Keywords

1 Introduction

The time frame for taking actions to effectively slow down further climate change is rapidly closing. Some of the changes in climate that have already come into effect do have increasingly noticeable impact and have increased the likelihood that extreme natural events will occur in the decades to come. Alongside climate mitigation, the adaptation to our new climate reality has become a task at hand. While mitigation strategies are concerned with influencing the occurrence of natural events, adaptation strategies focus on reducing the risk for these events to entail a disaster. Ideally, such an adaptation is sustainable, considering and supplementing mitigation efforts. On an international level, there are several frameworks and strategy papers aimed at reducing the risk for these extreme events to turn into disasters. One of them is the Sendai Framework for disaster Risk Reduction [1], adopted by the United Nations member states, another the recently published new strategy on adaptation to climate change of the EU [2]. In the past decade, many countries have adopted national adaptation plans (NAPs), for example Germany [3], or are in the process of doing so with support of the UNDP [4]. Those national plans serve as the framework for political decision making, but identify local authorities at the municipal level as the key actors to develop and realize adaptation measures.

The process for successful adaptation is advised to be an iterative one, where a number of steps are to be constantly repeated [5,6,7,8,9,10]. Essentially they can be summarized and described as:

  1. 1.

    Assessment of hazards and analysis of shocks: what events are to be expected in the area, in what likelihood, intensity and frequency.

  2. 2.

    Assessment of vulnerability, calculated as a median of three components: susceptibility to suffer damage in the event, the capacities to minimize negative effects of an incident and the capacity for long term adaptation and change.

  3. 3.

    Drafting and comparison of ideas for adaptation measures.

  4. 4.

    Development of a strategy how to implement the measures and implementation of that strategy.

  5. 5.

    Monitoring and evaluation of effects of the implementation of the strategy.

Individual cities are currently taking effort to survey their level of vulnerability and resilience and to draw up adaptation strategies [11, 12]. However, larger cities (with population of 500.000 and above) are by far the most active ones, while smaller cities seem to be over-challenged with the task [13, 14]. In Germany, only 14 out of the roughly 10.000 municipalities have a population greater than 500.000, housing about 17 percent of the country’s population [15]. To achieve resilience for a majority of the population, especially smaller cities should be supported in their efforts.

City planning is always about making compromises, as livability, sustainability and resilience are all multidimensional concepts that often have contrary demands [16, 17] and involve diverse stakeholders. This is now intensified by the pressure of climate change, especially since worldwide over 20% of cities will find themselves in climatic conditions that are unprecedented on the planet to date [18]. For those cities in particular, there are no historical lessons learned or empirical values on how to meet those demands. For many other cities, the experiences of cities previously located in such climate conditions that match the new climate reality of today can be a valuable source of information, if there are utilizable ways to harness this wealth of experience.

Some datasets on adaptation actions and insights are being shared and made publicly available, for example on the EU’s climate-ADAPT portal [19], the German portal for climate provision [20], the UN-Habitats Urban Resilience Hub [21] or CDPs open data portal [22]. There is great potential in accessing and using those and other data, for example coming from IoT devices or meteorological services like [23], when analysing vulnerabilities and opportunities, planning actions and measures and evaluating implementations. Recommendations to harness this potential are prominently published [2, 24, 25] but to date, data-exploration know-how is still mostly concentrated within a small group of experts and data-driven policy making is not yet integrated into culture within public administrations [24].

However, modelling and simulation methods which have been in use for engineering projects for decades, are increasingly gaining popularity within socio-ecological-economical contexts and policy making [25, 27,28,29]. They facilitate experimentation within a safe, virtual space before implementing changes in reality and can help to get a grip on the multiple and interdependent relationships within a complex system such as a city.

When models and simulations are implemented as a collaborative digital tool, for example as an interactive 3D map, they can aid all 5 steps of the adaptation process, from hazard and vulnerability assessment to monitoring and evaluation. They can enhance the transparency of decision making, improve communication between public administrations, citizens and other stakeholders and help to visualize the pros and cons of the effects that different policies will have under various scenarios. Explicit suggestions to use GIS-based software in urban adaptation contexts are numerous [1, 6, 26], but to date, this is not reality in the field.

GIS gain in importance, yet, their application is still comparatively heterogeneous and largely limited to inactual, insufficient resources of geodata not least due to a lack of standards and deficiency in network structure of geodata infrastructure at large (e.g. GDI-DE) and state-dependent (GDI-NI resp.). At the same time, there are also vast differences in the usage being determined mainly in the following two groups: well-structured and prosperous municipalities and cities utilizing GIS intensively, whereas smaller or fundless communities lack resources (staffing, hardware) to do so [27].

In this respect, we became aware of the whole-of-government approach that is “the movement from isolated silos in public administration to formal and informal networks (...), which are to be planned, implemented and evaluated with their participation, and the opportunities presented by the Internet to transform the way the government works for the people.” [28].

Hence, belonging to the seemingly small group of governmental officials who do possess know-how in (geo) data-exploration, we utilized our peculiar insider position and explored an open source grass roots approach and investigated: can we use already existing open source technology and open data to develop a low-threshold GIS-application that supports cities in their adaptation efforts? What are the characteristics of a system for diverse and complex city planning and decision making scenarios?

The remainder of the paper is structured as follows: First, we provide a short overview of related works and describe our product vision. We then compare indicator sets for the objective assessment of urban resilience and describe how to proceed from hazard maps towards a co-creatively workable digital twin in order to establish new networks to stack the systems of systems. We then lay out our concept for a resilience-platform which will be providing access to knowledge, before reaching our outlook and conclusion.

2 The Context of a Tool for Resilient Cities

We want to implement a tool that will allow stakeholders to identify and realize measures to meet urban needs and challenges on the way to becoming resilient cities. There exist a number of frameworks, definitions and guidelines for the assessment of resilience in cities. One widely accepted definition is the one published by OECD:

Resilient cities are cities that have the ability to absorb, recover and prepare for future shocks (economic, environmental, social & institutional). Resilient cities promote sustainable development, well-being and inclusive growth [29].

The concept of urban resilience is very closely related to that of urban sustainability. Indeed, we argue that resilience must be an integral part of sustainability when considered for a large time frame. Sustainability is defined as:

meeting the needs of the present without compromising the ability of future generations to meet their own needs [30].

When taking a systemic perspective, this entails a behaviour causing little or no damage to the environment, as it provides vital, indispensable services [31]. Arguably, when talking about sustainability of cities, the period of time implied is several decades, if not centuries. Within that period of time, extreme (natural) events will take place and a city can only be deemed sustainable if it is able to continue despite those events. Achieving sustainability only under unrealistic conditions, a fair weather sustainability so to speak, is in-consequent. The resilience of a city is therefore an integral part of being sustainable, because resilience is needed in order to be able to continue over a long period of time. Resilience is not evaluated with respect to a single event, but to all events that might be encountered within the period of time.

2.1 Product Vision

As stated in the introduction, it is our goal to facilitate urban climate adaption initiatives by consolidating existing frameworks. Bosetti et al. have reviewed 35 different frameworks for the assessment of risk, fragility and resilience [32]. They conclude that most risk and resilience frameworks are focused on stresses induced by exposure to natural hazards, but at that time few had been successfully validated at city level. Huovila et al. have performed a comparative analysis of standardized indicators for smart sustainable cities [33]. They recommend using standardized indicator sets as a starting point, but to select subsets specific to a city’s needs. Methods to look at the inter-dependencies and effects of actions across multiple indicator types are required, but still in the development phase and have to prove their applicability and usefulness in practice. This view is also present in [34], who argue to use meta-principles to guide system-level decisions in the transformation process of cities. Nieminen et al. [35] point out that in addition to an evaluation of the current state as an indispensable starting point, it can enhance decision-making to bring into focus options for change and drivers causing that change, especially when combined with dynamic models and simulations as well as social embedding.

We base our product vision on those recommendations for an interactive adaptation and decision-making process. In an internal design-thinking workshop, we refined the idea of basing the tool on an interactive 3D map, but this process was biased by the strong affinity of the participants for map-based visualisation. In order to sharpen the product vision further and include stakeholder requirements in a generalized form, we scanned the prominent guidelines and frameworks for the explicit use of the terms “map” and “GIS”. Additionally, we reviewed the publications of municipal central organisations [36,37,38,39,40]. The results are ordered by the 5 iterative steps of the adaptation process, as introduced in the first section of this paper.

For step 1 for example OECD [9] lists a “shock and stress map” as one of the valuable outputs of a resilience systems analysis, while [1, 6, 26, 41] refer to hazard maps as a means to engage and brief stakeholders for the assessment of hazards and analysis of shocks.

For step 2 the creation of risk maps is suggested by all of the above mentioned sources for assessment of vulnerability. In addition to serving as a static visualization of data, those maps should be interactive so that the perception of the stakeholders and local priorities can be incorporated, for example by performing a “mapping exercise” [6].

In step 3 during drafting and comparison of ideas for adaptation measures, maps are suggested to be integrated in planning [1], but their potential to be used as co-creation tools is not highlighted explicitly, contrary to step 2. In [36], tools for simulations are explicitly demanded, for example for the identification of cold air pathways.

Step 4 the development of a strategy how to implement the measures and implementation of that strategy is based on the preceding steps, but there is no explicit mention on how to utilize maps or a GIS in this process.

Step 5 for the monitoring and evaluation of the effects of the implementation of the strategy, maps are again mentioned by all of the sources.

This result affirms the recent findings of [42,43,44]: there is a rupture in the adaptation process, the changing of spatial scales is not seamless at the transitioning from step 2 to step 3. This is reflected in a change of tools: while assessment of hazards and vulnerabilities is done on a broader spatial resolution utilizing GIS, the concrete and practical measures are planned and implemented using only a very limited spacial scope, utilizing tools pooled under the term Building Information Modelling (BIM). The effects of the measures—step 5—are then again viewed and monitored on a larger spacial scope. Interoperability between BIM and GIS has to date not been established, but there are numerous and increasing efforts to achieve this [45].

To build a digital twin of the city is increasingly being explored as a solution that can integrate the suggested systemic approaches and bridge the gap between BIM and GIS [46]. The concept of digital twins emerged in engineering and manufacturing processes, and these origins influence the conception of digital twins for cities. As the concept of digital twins is adapted to different domains, fuzziness and a lack of consistency in the definition of what a digital twin is, has emerged [47]. We follow Jones et. al. in their definition and characterization of digital twins, emphasizing that a city digital twin is not simply a 3D model of the built environment, but rather a system of systems. However, we would like to point out that the time dynamics, especially the frequency with which the state of the physical twin and the digital twin are synchronised, should be explicitly addressed. In some research domains, near real-time coupling between the twins is stipulated [48], but in the domain of governmental geospatial data, the frequency of several years between data acquisition (for example airborne laser scanning) is not uncommon and there are no concepts on how to achieve real time measurements. When fusing this data of the built environment with more dynamic data, either by twinning other physical entities that make up a city such as vehicles, or by integrating weather data or energy consumption data, the question on coupling frequencies becomes interesting.

Shahat et al. as well as Mylonas et al. have recently performed a comprehensive review of digital city twins and have grouped the potentials of city twins into five categories: visualisation, planning and prediction, integration and collaboration, situational awareness and data management and [49, 50]. They found that very rarely human interaction or socio-ecological dynamics are integrated.

The European Commission has launched the “Destination Earth” initiative (DestinE) in 2021, which has the goal to build a digital twin of the entire planet. The project will start out by focusing on extreme natural disaster and climate change adaption at regional and national levels, refining to a local urban scale in a later phase after 2023 [51]. Thus, socio-ecological dynamics are not part of early project developement.

Simulation applications that do include living entities in spacial contexts are predominantly doing so in a 2D context [52]. While these dynamics might be of less importance in the assessment steps, we believe that they are to be a key element when drafting and implementing the adaptation measures, for example, when evaluating the percentage of residential properties located in high-risk zones or the annual number of residential properties flooded. To put it simply, there are two strategies to optimizing these resilience dimensions: either fortifying and installing protective measures to the area so it will not be flooded. Or motivating and empowering the people to either move into a low-risk area, or secure the financial resources to build back better after an extreme event.

The urban form has self-evidently been shown to influence resilience [53,54,55], but the effects that individual building projects have on the urban form are often not considered. Taking into account the urban form is technically possible, as spatial comparison operations already exist in common GIS tools [56] that enable the user to compare the similarity of a location with other areas using certain criteria and specifying the degree of similarity. As a city is a complex social-ecological-technological system (SETS), there are currently many additional approaches, such as comparing the topological similarities of urban road networks [57], or even micro-mobility patterns [58].

After this iteration of product vision refinement through literature review, a visual representation of our product vision is Fig. 1. With the whole-of-government approach, we elaborate on the seamless communication and engagement through effective networks and partnerships with other governmental and non-governmental entities.

Fig. 1
A model diagram represents the application, creation, and ideation which includes digit twin, decision making, data sources, forum, indicator values, interactive map, and communications.

Visualization of the product vision

Full size image

2.2 Comparing Indicator Sets

In a thorough, professional adaptation process key variables, dimensions and indicators relevant for the integration into planning processes have to be identified. Based on [32] and more recent developments in the field, we have selected three standardized indicator sets for closer investigation and possible integration into our tool:

  1. 1.

    The German Adaptation Strategy (DAS) [3].

  2. 2.

    The international standard ISO 37123:2019 “Inidicators for resilient cities” [59]. It is intended to be used in conjunction with the ISO 37120:2018—“Sustainable cities and communities—Indicators for city services and quality of life.”

  3. 3.

    The ETSI TS 103 463: “Key Performance Indicators for Sustainable Digital Multiservice Cities” [60], which is targeted at sustainability in the context of digitalisation and therefore contains some indicators, such as “digital literacy” that might not be deemed fitting to resilience adaptation.

These indicators shall be integrated into the objective functions of the simulations and decision support algorithms.

Out of the three selected indicator sets, two are explicitly designed for the use in city contexts, while the third takes data from a nation wide scope. The indicators are grouped into clusters in each set, see Table 1. These clusters differ in compass and underlying methodology, due to the different scopes of the frameworks. Taking a more detailed look at the indicators of each cluster, the overlap becomes larger. For example, the indicator “ Urban heat island ”, which is calculated as the maximum temperature difference between a point within the city and the surrounding area, is present in all three frameworks, but clustered into infrastructure, environment and climate change and climate resilience, respectively. Other clusters and their encompassed indicators are unique to a specific framework, for example Innovation. This cluster contains indicators such as accessibility of open data sets.

Table 1 Comparing indicator groups
Full size table

Combining all indicators from the three frameworks results in a set of 311 different indicators, with only 5 common to all three frameworks. This motivated us to prioritize the indicators for implementation. Since the ISO 37123:2019 is an internationally as well as nationally recognized standard and specifically derived for the evaluation of resilience in cities, we chose to give those indicators the highest priority. We then systematically grouped the indicators of ETSI TS 103 463 and DAS and identified those related to resilience. This lead to a combined set of 67 indicators. The indicators were mapped to the five steps of the adaptation process, in a similar manner as done suggested in the informative annex of ISO 37123:2019.

2.3 From Hazard Maps Towards a Co-creatively Workable Digital Twin

The integration of data into 3D maps from sources such as database tables, live sensor readings or aggregated data provided as a web map service (WMS), web feature service (WFS) or Spatio Temporal Asset Catalog (STAC) API is current state of the art. However, integrating all data sources needed for the calculation of the indicator set into one map and subsequently performing the calculations does require know-how in (geo) data-exploration. Seeking to lower this technological hurdle, we investigated which information needed for the computation and evaluation of the indicators is already openly available or could be made available from sources we possess in-house as a governmental agency.

One key information needed in the adaptation process is to be able to identify which areas are likely to be exposed to hazards, summarizing them under the term “high-risk areas”. While hazards can include “biological, environmental, geological, hydro-meteorological and technological processes and phenomena.” [59], there seems to be a focus on hydro-meteorological and geological ones. The information on hazards, such as their likelihood under different scenarios, their intensity and spatial extent can be found in a growing number of online sources. The DWD [23] provides information about extreme weather events such as extreme rainfall, extreme heat and urban heat island effects. The European Union hosts a geospatial data catalog [61] with map viewer, where currently about 1700 datasets can be accessed, including roughly 100 tagged with “climate”, another 45 “urban vulnerability”. The Federal Agency for Cartography and Geodesy (BKG) provides a map service where drought, heat, floods and heat events can be viewed [62]. The State Offices of Lower Saxony for Mining, Energy and Geology (LBEG) and the Ministry for Environment, Energy, Building and Climate Protection both provide a map viewing service where a larger number of layers have been integrated from various environmental, geological and hydro-meteorological sources for the state of Lower Saxony [63, 64]. The before mentioned DestinE initiative aims to further refine the predictions of hazards [51].

All these sources are great for a visual inspection and serve step 1, but they do not focus on an user interface to move past assessment of the risk imposed by the depicted hazards. While we are greatly looking forward to the open platform of DestinE, we see great potential and unmet demands in the facilitation of sharing experiences about adaptation actions.

2.4 Establishing New Networks to Stack the Systems of Systems

While individual efforts of cities increase to manage risks and crises, so are initiatives and the development of tools by smaller organizations or even state authorities. However, decisive investments in resources are too ubiquitous, which is why penetration of possibly valuable applications fails due to a lack of a critical mass of users. Nonetheless, what concerns the communal level, municipalities prove bilateral cooperations with neighbouring regions and extend these noticeably.

We have presented an iterative process for municipalities and cities to adapt to climate changes and become resilient: ideally along international standards such as ISO37123. Also, we have outlined how manifold and complex the concepts of both adaptation measures and resilience are. And so it is when striking a new path of product development that addresses the requirements of a tool of such societal relevance. We aim to overcome data and media disruption that facilitates the application of geodata for the special purpose of increasing the level of resilience. In the broader context of EU- and nationwide adaptation plans and among internationally applicable standards (i.e. ISO), we seek to develop a thorough understanding of users and user needs, becoming a leading multiplier to ensure resources for adequate resilience planning for any city and municipality, commencing in the German federal state of Lower Saxony. As a governmental entity that is envisioning a digital tool as described in the preceding chapters, we propose a new structure of co-creative partner and network management. When stepping in as a modern and sustainably acting organisation, whose aim it is to enable local municipalities of a state (Lower Saxony) and its cities to take action, we find this positioning mandatory.

A system is a network of interdependent components that work together to try to accomplish the aim of the system. A system must have an aim. Without the aim, there is no system. –W. Edwards Deming

Our forward-looking approach is more feasible to our general innovation initiatives as described in [65], where we currently build up new organizational structures: we constantly learn to recognize the strength of connecting and working in networks, in which we introduce and exploit new technological opportunities as well as focusing explicitly on user-oriented development. Thereby, we no longer acknowledge the ordinary provision of geodata up to date, but seek to supply servicing geodata in useful applications and web-services: here, the tool for resilience and climate change adaptation will be one first example. Critical for the implementation is our belief in the strength of reciprocity, organizing new ways of communication that are bound to the concept of solution-oriented feedback and feedforward [66].

Feedback is a communication instrument originating in cybernetics that is defined as “the transmission of evaluative or corrective information about an action, event, or process to the original or controlling source” [67]

Thereof, solution-focused feedback is a systemic approach actuating the efficient deployment of resources and concentrating on strengths, competence and the success of an outcome we are aiming at, rather than conventional performance evaluation [66].

Feedforward is defined as “the use of calculated or presumed future states of a process to provide criteria for its adjustment or control; anticipatory control” [68].

It is a modification of solution-focused feedback that is directed at specific potentials for change in the future and following the structure: (1) Target/Vision/desired Effect (2) Question (optional) (3) Desired behaviour. [66].

The information is ultimately linked to the further exploitation of content given that is ensured by occasion- and task-related communication activities and continually growing modes of cooperation and new partners.

In order to adhere to iterative adaptation processes, with these mechanisms in focus we will reliably validate and refine the selection of the indicator evaluation methods within the network and prove consistency and elaborate further on expanding the network management and structures.

2.5 Resilience-Platform: Providing Access to Knowledge

Within the current time window, while global digital twins are under development, we aim to put resources to use that already exist. The current design for the tool has a platform architecture at its core, as described in more detail in [65]. Diverse (micro)-services shall be combined to cover all 5 adaptation steps. Currently, a number of agile teams is working on those services at LGLN. They can be visualized as a pyramid, with cloud service supply at the base, enabling services like identity and access management in layer three, geodata infrastructure like a spatiotemporal asset catalog (STAC) or webGIS-as-a-service in the second layer and applications like the here described solution at the top, see Fig. 2.

Fig. 2
A pyramid model diagram represents the server that includes applications, geodata infrastructure, cloud service supply, and enabling services to the data management.

Service pyramid

Full size image

It will be a challenge to combine computationally complex operations (like simulating traffic, the effect of infrastructure and materials on floods or heat) into the platform and realize a pipeline that integrates the changes made to the modeled world while keeping previously made computational results and simulations as a reference. Nonetheless, implementations of those simulations do exist to date, hosted at the respective expert institutions. It will be our service to integrate those services that do provide an API or other externally available sources (WMS/...) directly into our platform and facilitate a data exchange with those that do not have APIs yet. We hope that through this curated process and bilateral communication, the accumulated feedback from users can help to speed up and qualitatively enhance the development of simulation tools and APIs.

It is essential that persons unfamiliar with GIS or technical subjects can interact intuitively with the system (stakeholder/decision makers/politicians). Otherwise they may be overtaxed in handling it or not be able to derive useful actions. Consequently it is necessary to reduce the complexity for the end user. This could be achieved by handling all operations and data acquisitions in the background that don’t need a direct interaction. For all data sources that are currently only manually accessible, like a city’s budget plan for example, a step-by-step-guide leading a user through all data sources that need direct interaction should be implemented. In addition, we want to become a service provider that plays an active part in adaptation processes, either by providing training, offering workshops or customization of digital services. Figure 3 shows a schema of the architecture concept. Since the computation of all indicators will require a larger number of data sources and simulations, we have included only a few examples into the graph for visualization.

Fig. 3
A flow diagram represents the application, creation, and ideation of external data sources such as weather, flood, health, and includes a scenario database, simulated scenarios, adaptation idea, and tool usage.

Graphical overview of tool architecture

Full size image

Like mentioned earlier, many parties struggle with the same problems in adaptation to new situations. To take up the example already mentioned, nearly all cities or local authority districts in Germany need to deal with heavy rain events and resulting floods. To ban the danger of “reinventing the wheel” and instead harness synergy effects and learn from others, the platform has to offer the possibility to share and showcase solutions, communicate lessons learned or ask experts in the network to improve ideas. To do so, the central 3D map with it’s possibilities to integrate and simulate data has to be extended by a project management system and searchable catalog.

A project can thus be enriched with metadata. These can, for example, contain a description of the project that provides an overview of the initial situation, actions and results. Due to the additional roles of users, which are created by a management system, it is possible to build a community within the system that can exchange ideas. On the one hand, it is possible to comment projects, which creates a dialogue about already existing actions. On the other hand, there is also the possibility of contacting the owner of a project in order to exchange information about measures and experiences in a more targeted manner.

Of course, the topic of resilience offers a large number of actions and topics that are not necessarily comparable, but all of them should find a place on the platform. As soon as the platform is accepted and used by a critical mass of users, projects from a wide variety of subject areas (see Table 1) are available. Searching through existing projects for suitable solutions to one’s own problems could quickly lead to frustration for the user if the search only shows projects that do not help in the present situation.

3 Conclusion and Outlook

While the short description of each project and topic labels will certainly give users a good starting point for finding relevant and interesting projects, we believe the tool would greatly benefit from a search and filter functionality. We want to investigate how the inclusion of similarity measures can enhance the search results. It is our hypothesis that through a comparison of the indicator values (multidimensional indicator vector) of different projects, the similarity of those values will indicate if a shared project’s adaptation measures have relevance for the user’s region of interest. Our second hypothesis is that the indicators alone will not suffice and should be complemented with geo-based similarity measures. It will be a focus of our future work to identify and integrate ways to infer the similarity of cities for adaptation processes.

In order to implement and test the hypotheses in the future, it must be possible to find areas that are similar in terms of content and geography. Additionally it is necessary to record changes in the indicator values after actions took place. It could be one of our users’ approach to say: is there a place that started out in similar conditions like we have now? What did they do to adapt? How big was the impact of that measure? So the platform needs to save the initial situation with its incoming data sources and the calculated indicator values for the project area. As soon as actions or changes are entered in a project, the indicator values must be recalculated and recorded as a multidimensional indicator vector. The initial situation should not be lost: it must be possible to compare the effect of actions with each other and to calculate which action is most effective. In addition, it should be possible to find projects that have a similar initial situation in order to derive actions from other projects.

The outlook on the imminent challenges encourages us strongly to adhere to new communication mechanisms based on effective, integrative feedback. We are eager to build up strong network connections as we have presented throughout in this paper.