Adaptive resilience of and through urban ecosystem services: a transdisciplinary approach to sustainability in Barcelona

. Ecosystem services (ES) from urban green and blue infrastructure (GBI) provide cities and their citizens with benefits necessary to cope with present and future sustainability challenges. Long-term comprehensive urban greening strategies, policies, and plans are thus central to the development of sustainable, liveable, and resilient cities. However, urban greening strategies are increasingly tailored to provide short-term benefits, overlooking the dynamic character of cities, which face both changes in the capacity of GBI to provide benefits (e.g., in the face of climate change) as well dynamic needs and preferences for benefits over time as a result of changing demographic compositions. Starting with a literature review on GBI-relevant policies for the city of Barcelona, we: (1) investigated the presence of resilience thinking in the city's GBI-relevant policies through the application of the urban ecosystem services resilience assessment matrix; (2) investigated resilience thinking in the city's policies through the co-development of scenario narratives of possible futures and their implications for ES; and (3) applied the narratives through a participatory approach to enhance stakeholder thinking on adaptive policies based on possible shifts in ES provision and needs. Application of the matrix identified two main gaps to current GBI-relevant policies related to two main aspects of resilience: recognition and assessment of possible future disturbances and changes, and low understanding of social and structural diversity. Through the co-development of four future scenario narratives (aging and shrinking population, enhanced tourism, gender inequalities, and global warming), stakeholders identified the most susceptible ES in the city of Barcelona. Workshop participants indicated mental well-being, regulation of microclimate, social cohesion, air purification, physical recreation, runoff control, and soil permeability as ES with the widest capacity–demand mismatch. The results elicited discussion around GBI and ES resilience, addressing the need for intersectoral policy integration (including housing, education, and mobility) and for fostering a wider understanding of the role of institutions in providing for a resilient urban future. Through the use of scenario narratives, and highlighting the potential of co-creation, the proposed approach enhances critical thought around ES resilience among key players in the city. The study thereby supports the development of a comprehensive resilience strategy for Barcelona and indicates pathways for how other cities can change their current urban trajectory towards sustained ES flows.


INTRODUCTION
With urbanization on the rise (Güneralp et al. 2017, UN 2019, ecosystem services (ES) provided by urban green and blue infrastructure (GBI) offer multiple health and well-being benefits necessary to cope with present and future urban challenges (Gómez-Baggethun and Barton 2013, Gascon et al. 2015, van den Berg et al. 2015. These ES include, for example, microclimate regulation, runoff control, and opportunities for outdoor recreation (Gómez-Baggethun and Barton 2013, Haase et al. 2014). Long-term urban policies and strategies can play a central role in maintaining and increasing ES to create more sustainable, liveable, and resilient cities (Ahern et al. 2014). However, current urban strategies often overlook the sectoral integration that is needed to harness the potential of GBI to support sustainable urban transformation and create resilience around human wellbeing (McPhearson et al. 2016). Moreover, to maintain well-being related benefits over time, cities need to build capacity to adapt both to external drivers of change, such as global climate change and pandemics, as well as to inherent changes, such as shifting demographic composition (e.g., due to an aging population or migration) and related changing demands for ES benefits from GBI over time.
The need to build urban resilience has increasingly gained attention in the last decade in both science and practice (e.g., Meerow et al. 2004, Quinlan et al. 2016, Moser et al. 2019. Resilience studies help us to understand complex social-ecological systems and how to plan and manage them for sustainability, not least with respect to climate change (Elmqvist et al. 2019). However, urban resilience is generally vaguely defined, which makes its application difficult as an analytical framework (Sellberg et al. 2018). Resilience is often interpreted as being positive, which is misleading. In response, Elmqvist et al. (2019) introduced a more precise understanding of urban resilience connected to sustainable transformations of urban areas. Whereas urban sustainability and ES are considered a normative concept representing a "positive" vision for the future of the society (Schröter et al. 2014, Romero-Lankao et al. 2016, the authors describe resilience as a neutral, non-normative concept (e.g., Ernstson 2013) and an intrinsic property of social-ecological urban systems. To fit with a system in constant change, resilience is defined as the "capacity of an urban system to absorb disturbances, reorganize and maintain essentially the same functions over time and continue to develop along a particular trajectory" (Elmqvist et al. 2019:269). This interpretation suggests that resilience can be a barrier to the desired transformations if the city is following a pathway that is not sustainable, and that it needs to be reduced to enable transformation (Moore et al. 2018) toward sustainability. In framing these considerations, we position the GBI potential of generating multiple ES as the core feature to be sustained. We argue that this more dynamic definition will help to secure human well-being in the face of challenges related to climate change and social transformations.
Securing a resilient ES flow in cities has raised substantial policy interest (Gómez-Baggethun and Barton 2013, McPhearson et al. 2015, Elmqvist et al. 2017. Nevertheless, tailored policies and interventions are needed to reduce the resilience of barriers to equitable access to ES . Moreover, tailored policies should build and increase resilience around the factors that enable the flow of ES benefits over time (McPhearson et al. 2015). In light of diverse and changing demands for ES in the future , planners and decision makers must address uncertainties when designing adaptive policies. As Walker et al. (2001) suggest, policies should be "devised not to be optimal for a best estimate future, but robust across a range of plausible futures." However, from a practical perspective, it is far from obvious what these barriers and enabling factors are, and how to engage with their resilience.
Transforming or guiding cities toward desired and sustainable futures, in which ES capacity aligns with ES demands (Villamagna, et al. 2013, Baró et al. 2016) requires improved integration of resilience thinking into urban policies. Here, we address three specific questions for guiding ES resilience-oriented policy-making: "resilience of what?", "resilience to what?", and "resilience for whom?" This strategy implies a broader understanding of: (1) external (e.g., climate change) and inherent drivers of change (e.g., demographic changes; i.e., "resilience to what?"); (2) whether and how these drivers and changes affect the generation of urban ES and the realization of their benefits (i.e., "resilience of what?"); and (3) who has access to ES considering evolving ES demands over time (i.e., "resilience for whom?").

CONCEPTUAL APPROACH
To govern the flow of ES benefits effectively, urban policies must acknowledge urban ES benefits as fundamentally co-produced by natural and human assets (Ernstson 2013, Langemeyer and. We understand GBI as the source of local ES, which are negotiated, regulated, and distributed across urban social-ecological systems (Andersson et al. 2019). The wider social-ecological system is, in turn, strongly influential in shaping and maintaining the quality and functionality of GBI. To gain a systemic understanding of GBI and the availability, accessibility, and fair distribution of ES, Andersson et al. (2019) proposed a framework of three interconnected systemic filters: infrastructure, institutions, and perceptions. These filters are recognized factors that affect the capacity of GBI to produce ES and hinder or facilitate the realization of ES benefits by different beneficiaries. Although GBI is critical to guarantee the capacity of ES, ES are realized in the complex interplay of grey infrastructure (different types of housing developments, transportation networks, etc.), actors, roles, rights, responsibility, and management (institutions), as well as specific needs, knowledge, practice, and identities (perceptions; Andersson et al. 2019).
We combined Andersson et al.'s (2019) three interconnected filters approach (infrastructure, institutions, and perceptions) with seven ES resilience principles proposed by Biggs et al. (2012): (1) maintain diversity and redundancy, (2) manage connectivity, (3) manage slow variables and feedbacks, (4) foster an understanding of complex adaptive systems, (5) encourage learning and experimentation, (6) broaden participation, and (7) promote polycentric governance systems. We argue that adapting the ES resilience principles to the urban realm could support policymaking in pursuing resilience around the flow of ES by anticipating and systemically adapting urban social-ecological systems to different drivers of change (see the urban ecosystem services resilience assessment matrix in Appendix 1). The (assumed) objective of urban policies would be to sustain and maintain both currently used and potential but latent GBI benefits over time. We believe that by applying and adapting Biggs et al.'s (2012) seven principles to the three filters in an urban context, policies (as an institutional tool) can shape infrastructures to ensure their maintenance and connections (principles 1 and 2). Also, translating into practice principles 3 and 4, through constant monitoring of slow variables and feedback interactions among all three filters, would support policy adjustments and adaptations to multiple drivers of change. Although policies can reframe the beneficiaries' perceptions and strengthen their capacity to benefit from the system, guiding the system through changes (principle 5), beneficiaries' perceptions can also influence infrastructure through GBI co-design and as well bottom-up influences on GBI governance (principles 6 and 7; Barnaud et al. 2018). Departing from the premises of this conceptual framework (Fig. 1), we address four research objectives, as described below, to trigger policy adaptations that enable a resilient ES flow.
Integrating ES resilience principles into policies and planning approaches (institutions) can function as a lens for identifying leverage points for unlocking the flow of ES from nature to humans, under both current and potential future conditions Kronenberg 2018, Elmqvist et al. 2019). Thus, the first objective of our empirical approach is to address ES resilience within GBI policies and question whether policies can ensure the generation of urban ES and the subsequent realization of their benefits over time ("resilience of what?").
To avoid narrowly focusing on single external drivers of change (e.g., climate change) in policy development and definition, our second research objective was to co-develop scenario narratives to trigger multilayered resilience thinking (Schewenius et al. 2014, Wiese 2016. This objective is meant to support the integration of different drivers of change into policy-making, including their interactions, feedbacks, and combined effects on the capacity and demand of ES ("resilience to what?").
Based on the assumption that GBI benefits tend to be distributed unequally among different social groups (e.g., Ernstson 2013, Baró et al. 2019, we assume that building resilience around a trajectory of sustainable development needs to be inclusive (Tozer et al. 2020). Therefore, we need to understand not only which ES would be more susceptible to changes under different future scenarios, but also which social groups might become excluded from the future ES flow of benefits. The third objective is to develop a participatory approach to understand the potential  (perceptions, institutions, infrastructures) and ecosystem services (ES) resilience principles (P1 to P7) that affect the capacity of urban green and blue infrastructure (GBI) to produce ES and unlock the flow of benefits, under different drivers of changes. The study framework is built on Biggs et al. (2012) and Andersson et al. (2019), using the ES resilience principles and filters as pathways for sustainable policy-making. Arrows from left to right suggest that policy-making might work through the systemic filters and resilience principles to ensure resilient ES flow and counteract drivers of change that undermine the uptake of GBI benefits. changes in ES capacity (infrastructure), citizens' desires or needs (perceptions), and the future distribution of benefits among different social groups ("resilience for whom?").
Our fourth objective is to analyze the proposed policy measures and their integration of resilience thinking into such policies. Navigating the complexity of assessing future scenarios into participatory resilience thinking will lead to the identification of adaptive policy measures for building resilience around the future flow of urban ES, taking into consideration their capacity and demand. This objective will allow us to assess the proposed conceptual and participatory framework for fostering resilience thinking around ES in urban systems.

CASE STUDY
Our empirical study focused on the city of Barcelona. With 1.62 million inhabitants, Barcelona is the second largest city in Spain and one of the most densely populated cities in Europe, with approximately 16,000 people/km² (Barcelona City Council Statistical Yearbook 2019). In recent years, the Barcelona city council has embraced an ES-based approach to urban greening policy, developing several strategies and plans to support the city's trajectory toward a more sustainable future (these plans include the Barcelona Green Infrastructure and Biodiversity Plan 2020 through the coordinated work of several city departments including the city's Urban Resilience Department. Our study is embedded in the current discourse around resilience building and is intended to assess urban ES resilience and explore new policy options to secure and unlock the future flow of ES. It builds upon a workshop that was part of an ongoing stakeholder engagement process organized by the Institute of Environmental Science and Technology (ICTA) in Barcelona since 2013 (in the context of the European research projects FP7-Openness, BiodivERsA3-ENABLE, EC-H2020-NATURVATION). The stakeholder engagement process was initiated to promote new relations, knowledge sharing, and empowerment among GBI-related stakeholders in Barcelona. The process creates an interface of current urban policies with research, as well as civic and private initiatives, touching upon themes such as public health, social inclusivity, and just planning in urban greening.

METHODS
We adopted a transdisciplinary and mixed-method research approach (Turnhout 2019) of sequential steps inspired by participatory multi-criteria evaluation approaches (e.g., Langemeyer et al. 2018) to address urban ES resilience. First, we developed an urban ecosystem services resilience assessment matrix to perform a systematic analysis of how municipal sustainability policies aligned with adapted ES resilience principles (following Biggs et al. 2012, Borgström et al. 2015, Nykvist et al. 2017, Andersson et al. 2019. Second, we co-created four scenario narratives to investigate potential external drivers of or inherent changes in the social-ecological system (Nelson et Biggs et al. 2012, Borgström et al. 2015, and Nykvist et al. 2017 al. 2005). The themes for the scenarios were defined through multiple iterations with the city's Urban Resilience Department. Third, the narratives were used in an expert stakeholder workshop to trigger critical thinking about possible future shifts in capacity and demand of ES. Finally, this new understanding was used to prompt expert stakeholders to reflect on the adaptation required for a transition to more desirable futures and to propose targeted policy interventions.

Policy analysis
We limited the analysis to policies developed at the city level and those that were relevant to the urban GBI and related ES development that dealt with sustainability, climate, and greening. Ten GBI-related policies were screened according to their potential alignment with and treatment of urban ES and their relevant resilience aspects (Appendix 2). The process followed a two-step approach. The first step included the screening of all 10 policies, with two objectives: (1) to translate the ES resilience principles into a context-relevant articulation that could inform an understanding of "resilience of what and to what?", and (2) to verify and assess the relevance and scope of the selected policies relative to the ES resilience framework. Consistent with Geneletti and Zardo (2016) and Rozas-Vásquez et al. (2018), our policy screening did not employ a strict keyword-based content analysis, but relied on both explicit and non-explicit qualitative content analysis applied to all sustainability-related city policies. Neither ES nor the resilience terminologies are yet standardized, and both are contingent on the context and must be sensitive to alternative languages (e.g., Camps-Calvet et al. , Meerow et al. 2016, Sellberg et al. 2018. Thus, in the first screening step, we translated the generic ES resilience principles identified by Biggs et al. (2012) to a case-relevant set of variables described in a language that resonated with how the policies were formulated.
We then performed a full analysis based on the urban ecosystem services resilience assessment matrix (Table 1). This matrix builds on Biggs et al. (2012), Borgström et al. (2015), Nykvist et al. (2017), and Andersson et al. (2019), and aims at adapting the ES resilience principles to the urban realm. Assessment variables relating to the ES resilience principles spanned sociocultural and biological diversity, urban morphology, planning approaches and normative context, consideration of external drivers (tourism, climate change, housing, technological innovation, demographic and political change), and inherent changes (human preferences and lifestyle) in the urban social-ecological system (Table 1). We translated qualitative judgments (high, medium, and low) https://www.ecologyandsociety.org/vol26/iss4/art38/ regarding the level of incorporation of the ES resilience principles into each policy using a score from 1 (low incorporation) to 5 (high incorporation) to facilitate the representation of the results. The connections between the variables and the seven ES resilience principles are detailed in Appendix 1.

Co-development of scenario narratives
The development of scenario narratives as input to the expert stakeholder workshop was done in close collaboration with the city's Urban Resilience Department in spring 2019. Scenario narratives describe plausible futures affecting Barcelona's socialecological systems configuration (Palomo et al. 2011, Priess andHauck 2014). The co-development of scenario narratives served, in particular, to identify and highlight critical external drivers of change and inherent system changes with potential negative effects on resilient ES flow, such as mismatches of ES capacity and demand (Villamagna et al. 2013, Baró et al. 2016. The following nine ES were deemed relevant for purposes of this study: regulation of microclimate, runoff control, air purification, carbon sequestration, noise reduction, social cohesion, physical recreation, mental well-being, and tourism recreation. This selection of ES, which was adapted through discussions with the Urban Resilience Department, built on results from a preceding workshop with local stakeholders that focused on the prioritization of urban land uses for the local production of ES ).

Participatory workshop
The participatory workshop underlying this study took place in Barcelona on 6 June 2019 and involved 49 "expert stakeholders" (including the organizers) from 27 different organizations, including local nongovernmental organizations, private consultancies, small enterprises, different levels and departments of public administrations, and academia (Appendix 3). Following a general introduction of the results of the policy analysis, participants were divided into four heterogeneous breakout groups to work on two main exercises: simulation of shifting ES capacity and demand, and development of policy options to build resilience around ES. Each of the four groups was assigned to work with one specific scenario narrative, which situated the discussion on shifting ES capacity and demand. For each future scenario, participants were asked to assess potential changes in ES. Changes were assessed based on either: higher or lower number of users and relative awareness of benefits leading to increased or decreased pressure on urban GBI (indicating a shift in ES demand), or higher or lower availability of urban GBI leading to increased or decreased ES capacity. In the analysis of the workshop outcomes, numerical values were assigned to determine whether the assessed demand for and capacity of the single ES in a given scenario would: decrease substantially (−2), decrease moderately (−1), stay unchanged (0), increase moderately (+1) or increase substantially (+2). In presenting the results, we use average values, but we acknowledge that considering the variability in responses may allow detection of a lack of consensus over changes in capacity and demand of ES and may be useful as an indicator of uncertainty (Langemeyer et al. 2018).
Based on the results of the ES capacity and demand assessment in the different scenarios, participants were asked to propose tailored and adapted policy interventions to build resilience around ES flows, tackling the specific issues arising from the different scenarios. The policies and interventions proposed were then clustered into different policy sectors together with the workshop participants, and proposed policy measures were further analyzed by the authors after the workshop. For each policy option proposed, the analysis identified which of the three systemic filters (infrastructure, institutions, and perceptions; Andersson et al. 2019) was primarily addressed, as well as which type of ES resilience principles were incorporated using the urban ecosystem services resilience assessment matrix (Table 1). This process allowed us to assess whether the applied resilience thinking approach was reflected in the policy measures proposed. Specific modalities of each workshop session are detailed in Appendix 4 and Appendix 5.  (Cortinovis andGeneletti 2018, de Luca et al. 2021), these urban policies primarily refer to regulating and cultural ES. To some extent, the policies also incorporated the ES resilience principles, which we assume will help to sustain ES in the future. Our analysis considered the degree of integration and consideration of the seven principles of ES resilience within four core policies (Fig. 2). We found the infrastructure filter, or the structural role of GBI in the city, was broadly addressed in terms of both the current state of GBI and future actions to improve the flow of ES (principles 1 and 2). For instance, the Climate Action Plan 2030 sets the goal to increase urban green space by 1.6 km 2 (equivalent to an additional 1 m²/inhabitant), which corresponds to an increase of 15% of the city's urban green areas by 2030, an ambitious objective in a compact city such as Barcelona.

Ecosystem services resilience in existing policies
Biological diversity and redundancy (principles 1 and 4) are specifically accounted for in terms of biodiversity preservation in the Barcelona Green Infrastructure and Biodiversity Plan 2020, in the stimulus program for the city's urban green infrastructure, and in the Trees for Life Master Plan. However, structural and socioeconomic diversity of the urban area are not addressed, except for in the Climate Action Plan 2030, which considers some socioeconomic issues along with demographic variables (principles 1 and 4). These issues include inherent changes such https://www.ecologyandsociety.org/vol26/iss4/art38/

Fig. 2.
Results from the policy analysis using the urban ecosystem services resilience matrix. The integration of the ecosystem services (ES) resilience principles (P1 to P7) into each green infrastructure policy ranges from 1 (low incorporation) to 5 (high incorporation). SES = socialecological system; CAS = complex adaptive system as possible increases in population, migration, and external drivers such as the effects of climate change on vulnerable neighborhoods and populations. The Climate Action Plan 2030 also refers to other relevant policies and plans, clearly showing links, connections, synergies, and opportunities with other policy sectors, acknowledging institutional flexibility and a systemic approach (principles 3, 4, and 5). None of the other three plans account for possible shifts in future ES demands, thus neglecting the perceptions filter, at least in the face of systemic changes (principles 3 and 4). Furthermore, the policy analysis indicated an explicit but one-dimensional focus on adaptation to climate change and general lack of attention to the management of other disturbances and drivers of change (e.g., climate, demographic, economic, political, technological innovation, human preferences and perceptions, tourism, housing, and land-use planning; principles 3 and 4). Greening strategies did recognize major disturbances caused by plagues and climate-related events, but did not consider other possible changes and disturbances (principles 3 and 4), such as those related to growing and shifting ES demands or capacities.

Scenarios of change affecting ecosystem services resilience
After several iterations with the Urban Resilience Department, four scenarios were proposed as the most relevant drivers of change to Barcelona's resilient ES flow: an aging and shrinking population, enhanced tourism, gender inequalities, and global warming ( Table 2, Appendix 6). The four scenarios and the results of the policy analysis constituted the entry points for triggering participants' thoughts on possible disturbances and changes in capacity and demand sides of critical ES in the future during the stakeholder workshop. It is worth highlighting that the four scenarios were defined in spring 2019 before the global COVID-19 pandemic and the ensuing tourism crisis.

Future ecosystem services capacity-demand gaps and potential policy responses
Shifting ecosystem services capacity and demand During the participatory workshop, stakeholders were asked to assess and discuss the changes in capacity and demand for each ES to explore the potential gap between the growing demand for and recessing capacity of given ES (Fig. 3). Across the four breakout groups working on the different scenarios, participants generally assessed that the ES demand will remain stable or increase, whereas the capacity of ES will remain unchanged or decrease. The discussions centered around ES that were deemed critical for urban sustainability and citizen's well-being, including regulating ES such as microclimate regulation, air pollution, and runoff control (Larondelle et al. 2014), and cultural ES such as mental well-being, physical recreation, and social cohesion (Andersson et al. 2015). The workshop revealed that mental wellbeing is the ES most susceptible to future changes, indicated by the capacity-demand gap (Fig. 3), followed by microclimate regulation, social cohesion, air purification, physical recreation, and runoff control and soil permeability. In contrast, noise Table 2. Four scenario narratives co-developed with the Urban Resilience Department. Each scenario is centered around a single narrative that represents a major challenge for urban sustainability.

Scenario #1 Aging and shrinking population
There is an outmigration of the young population and ageing of the resident population. The pressures on health, mobility, housing, job availability, and social services have increased substantially. The resulting effects are depression and loneliness in elderly people, a lack of opportunities to engage (socially and economically), and in public health and well-being. The city lacks accessible open green spaces which have the capacity to provide manifold social and environmental benefits.

Scenario #2 Enhanced tourism
Mass tourism is a source of wealth but also of complex challenges. The pressures on housing, services, and urban space availability as well as on the urban environment have substantially increased. The effects include rising prices, increase in illegal activities, overcrowded and degraded open spaces, and changing attitudes of residents living in affected central neighborhoods.
Affected residents are deprived of available green spaces and beaches for recreation. Several local movements have emerged and taken action.

Scenario #3 Gender inequalities
An increasing number of women are reporting negative experiences from their visits to open spaces, also due to misperception and disinformation. Public spaces are dominantly used by men due to changes in users' perceptions. Equal access to green and open public spaces is questioned. Women's access to green spaces is limited, as they perceive them to be unsafe. The female population is deprived of benefits related to urban nature.

Scenario #4 Global warming
Climate change has intensified and is affecting the city and its residents. Increase in the number of torrid days, droughts, fires, rain-shortages, and water scarcity presents an immense challenge for the future of the city. Some residents are affected more than others, particularly vulnerable groups such as the elderly, children, and pregnant woman. Similarly, heavily built up areas are affected the most. An increased importance of friendly and climate-resilient open spaces is highlighted.
reduction, tourism recreation, and carbon sequestration were estimated to remain relatively stable in terms of their capacity and demand. Carbon sequestration was considered to be the least important ES provided by urban GBI. Tourism recreation had the smallest gap between capacity and demand. However, participants agreed that mass tourism and the related demands for tourism recreation generated by GBI will affect the availability of and accessibility to green spaces. In turn, this situation would negatively affect the provision of other GBI-related benefits, including mental and physical recreation and social cohesion, all of which are derived from GBI through direct nature experiences (Bratman et al. 2019) and which appear to be more vulnerable to changes. Given the complexity of the resilience thinking exercise, consensus was difficult to reach in some cases, especially for assessing ES capacity. For instance, the group addressing the global warming scenario had diverging views on changes in capacity of regulatory services and gained consensus on "no changes" only after assuming that the current policies in place, including the Climate Action Plan, Green Infrastructure and Biodiversity Plan, and Tree Master Plan, will improve ES capacity. The group addressing the gender equality scenario did not arrive at a consensus over changes in capacity of social cohesion, mental well-being, and related relational benefits and recreational opportunities. Some participants were of the opinion that ES capacity will increase simply because women are excluded from daily use of the public space, whereas others opposed this idea. For the enhanced tourism scenario, some participants suggested that the ES capacity of recreational opportunities, tourism, and economic benefits will increase only in core urban areas because of new developments and emerging economic opportunities, whereas ES capacity will decrease in the city's outskirts.

Adaptive policies to maintain ecosystem services potential
To match an overall increased demand and uncertain capacity of ES, the policy measures proposed by participants can be summarized in two larger clusters. The first proposed policy measure was to include measures aimed at increasing and sustaining current GBI and the related capacity to generate regulating ES and cultural ES over time. The second was to include measures that would improve access to GBI for city inhabitants through a more inclusive and participatory design that would mitigate pressure on GBI, primarily through limiting the number of visitors, especially tourists. We summarized the linkages between the identified policy clusters, the three filters, the ES resilience principles, and the capacity and demand of cultural ES and regulating ES (Fig. 4).
Most of the greening measures proposed in the workshop included direct interventions in the current GBI and built infrastructure of the city, such as expanding GBI and further integrating green and grey infrastructures, and improving the quality of GBI by making it more socially and ecologically interconnected. It was suggested that greening and climate measures promote diversity and redundancy and increase the extent of GBI (Fig. 4). Suggested actions ranged from creating new green areas, increasing vegetation and biodiversity, developing sustainable urban drainage systems and permeable surfaces, and improving nodes and connectivity among different green areas. The measures suggested for mobility primarily focused on enhancing connectivity (principle 2) by means of redesigning sustainable and slow mobility to leave space for green areas, and improved integration of GBI and grey infrastructure. Also discussed was the importance of rethinking the role of the airport in the city, and limiting its further expansion.
Many of the policy interventions that were discussed target GBI through processes of maintenance and management rather than addressing GBI directly. Examples included involving citizens in urban transformation processes and projects to boost community care and maintenance of the GBI, as well as promoting relational and well-being aspects such as managing urban gardens. Urban planning is one of the guiding instruments at the core of institutions (Andersson et al. 2019), and many measures proposed refer to land-use rights, actor roles and responsibilities, and the different ways people can be involved in changing how land is planned for and used. Measures related to institutions and governance processes covered a range of options, such as a proposal to include technological innovations and greening https://www.ecologyandsociety.org/vol26/iss4/art38/ Fig. 3. Assessment of the ecosystem services (ES) capacity and demand gaps in the four scenarios (S1 to S4). The darker shading indicates a decreasing trend in ES capacity (-1 = moderate decrease; -2 = substantial decrease), whereas the lighter shading indicates increasing demand for ES (+1 = moderate increase; +2 = substantial increase). The colored bars are averaged values from the ES assessment exercise, which highlight the majority of the respondents' opinions when consensus was not reached. Values equal to zero correspond to no changes in the ES capacity and demand assessment.
criteria as a requirement in building codes, mostly acting on regulating and cultural ES capacity (principles 1 and 2), but also increasing accessibility and inclusivity of existing green areas, including the security of those spaces (i.e., addressing cultural ES demand). A number of measures addressed the need to manage tourist flows so as to not overcrowd urban green areas, and to limit cultural ES demand in specific places in the city and its surrounding areas. Proposals addressing tourism included financial adjustments and suggestions to redirect tourism taxes to protection, maintenance, and improvement of green areas. Other proposals touched on greening, health, and including access to green areas in the annual health report as potential benefits for the human health of the city and managing slow variables and feedback (principle 3).
Another suggestion for the scenarios was to include the elderly and women in community initiatives for improving and maintaining local GBI (principles 1 and 6). Workshop participants also proposed measures for strengthening participation of vulnerable groups in GBI co-design, which would reframe both the physical space and how the urban environment is perceived (principle 6). Participants addressed perceptions in several other measures, which differed slightly from the respective scenarios. For example, urban farming was mainly addressed in the aging scenario and was recognized as a main factor enabling GBI benefits for the elderly, which confirms findings by Camps-Calvet et al. (2016). In contrast, the gender inequalities scenario generated more attention to security (principles 1 and 6), improved accessibility (principles 1 and 6), and more inclusive codesign, co-management, and co-maintenance (principles 4, 5, 6, and 7). Feelings of insecurity and the unjust distribution of cultural ES benefits among different citizens groups were at the core of this discussion, which aligns with Maruthaveeran and van den Bosch (2014). Education (learning) and awareness-raising measures (principle 5) were mentioned as critical for involving different target groups (elderly, children, women, students) and mostly refer to environmental education activities.

Advancing multifaceted urban ecosystem resilience assessments
The value of applying the urban ecosystem services resilience assessment matrix lies in the translation of theory into an analytical and operational framework for which we can assess urban policies, embracing the complexity of cities as adaptive social-ecological systems while helping to make them manageable (Table 1). Recognizing and connecting relevant system components is critical for building resilience (Biggs et al. 2012), and problem framing may serve as an indication of what solutions are being sought (e.g., Dennis and Brondizio 2020). For instance, applying the urban ecosystem services resilience assessment matrix to the sustainability policies of the city of Barcelona revealed a sophisticated incorporation of principles 1 and 2 mainly related to physical connectivity and consideration of biological diversity. Incorporating these principles would enhance the condition, accessibility, and connectivity of the urban GBI, connecting biological diversity and also partly including the social and structural diversity of the urban GBI. Principles 6 and 7 are well embedded in current policies, indicating the city's growing interest in actively engaging its citizens and stakeholders in decision making processes and defining responsibilities, tools, and collaboration methods. Further, we found that scientific and technical knowledge of urban ES and GBI (principle 5) is well integrated into all of the policies, providing an entry point for science-driven approaches for addressing the existing gaps. Despite progressive incorporation of several crucial principles, other crucial aspects of urban diversity, such as social and structural diversity, are not properly addressed in the policies we assessed. Furthermore, none of the policies we analyzed had a clearly defined approach for identifying emergent signals, disturbances, or unforeseen changes, nor for potential respective responses (principles 3 and 4) Our study confirms observations by McPhearson et al. (2015) that urban ES are particularly valued in the context of climate adaptation, improved citizens' health and well-being, and as means for enhancing the city's resilience (resilience through urban ES). However, the resilience of urban ES ("resilience of what?") under changing conditions ("resilience to what?"), and possible changing demand ("resilience for whom?") (i.e., possible variables, disturbances, slow feedback, and changes) have yet to be adequately considered. Although the policies consider the potential effects of climate change on current GBI, application of the urban ecosystem services resilience assessment matrix can support a multilayered resilience assessment to tackle vulnerabilities to multiple external (e.g., pandemic, technological innovation) and inherent (e.g., demographic, lifestyle changes, housing, and socioeconomic) drivers of change.

Fostering resilience thinking: to what, of what, for whom?
Departing from the gaps identified in the policy assessment, and to elicit answers around our three main research questions ("resilience of what, to what, and for whom?"), the workshop participants were tasked with co-developing four future scenarios. The scenario exercise is at the interface of science and policy and in itself is a step to furthering resilience thinking (Pereira et al. 2019). After familiarizing themselves with the different future scenarios ("resilience to what?"), participants generally found it easiest to start with evaluating ES demands. Demand for cultural ES (i.e., citizen's needs) was assigned the highest importance and was the focus of most discussions, which is in line with the majority of urban ES studies that focus on (changing) ES capacity (Haase et al. 2014). Perceptions from a diversity of beneficiaries were explored (i.e., elderly, residents and tourists, and women) as well as potential changes in ES demand from vulnerable groups. Equal access and inclusively designed green spaces are considered crucial to address shifting ES demand in the future (e.g., Ode Sang et al. 2016, Fumagalli et al. 2020, not least to foster civic stewardship (Langemeyer et al. 2018, Andersson et al. 2019). Shifting ES demands and beneficiaries' perceptions were not prominent in the analyzed policies, suggesting that the codevelopment of the scenario on aging, gender, and tourism strongly supported resilience thinking toward this point ("resilience for whom?"). Although participants were instructed to work with a single scenario, they tended to consider different drivers, particularly mass tourism in conjunction with global warming. This tendency indicates the potential of stakeholder workshops for examining difficult problems of complex interacting external and inherent changes in relation to multiple ES demands, which can be difficult to capture through other approaches (e.g., computational modeling; Pereira et al. 2019, Scolozzi et al. 2019). Further, on the perception filter, the use of Ecology and Society 26(4): 38 https://www.ecologyandsociety.org/vol26/iss4/art38/ the scenario narratives helped participants to understand the potential changing needs of different groups of GBI users, and subsequently provided an improved understanding of current and future demand for ES. Scenario narratives supported not only stakeholders' understanding of the consequences of possible development paths ("resilience of what and to what?"; Nakicenovic and Zimm 2017), but also contributed to developing an inclusive vision for future sustainability ("resilience for whom?") and to proposing concrete policy adaptations for achieving sustainability (Palomo et al. 2011).

Transdisciplinary co-creation toward adaptive sustainability policies
The adaptive policy measures suggested during the workshop for tackling future ES capacity-demand mismatches embrace an understanding of the interconnected character of grey and green infrastructure in the city. The GBI and their different ecological qualities provide the first necessary precondition for ES (Andersson et al. 2019), including their maintenance and resilience over time. These strategies can be understood as promoting both the diversity of GBI (principle 1) and its connectivity (principle 2), which are two central aspects of ES resilience. Interestingly, several policy suggestions addressed transport and mobility, thus parts of the grey infrastructure, a sector that strongly affects the availability, or lack of, GBI benefits (Biernacka and Kronenberg 2019). In addition to making the structure of the GBI more resilient, the effects of human activity and the modification of urban ecosystems depend on good management and long-term governance to maintain or, in many cases, strengthen their ecological qualities. Interestingly, Amorim Maia et al. (2020) found in a recent study from Barcelona that aesthetically less-pleasing green spaces (those lacking monuments, fountains, with low quality of common GBI indicators) could lend to improved inclusivity and foster social inclusion, while mitigating social segregation. When asked to develop policy adaptations based on the identified shifts in ES capacity and demands, participants typically developed measures that touched on policy sectors such as mobility, tourism, and health.
Although it is widely recognized that urban planning decisions in sectors such as land use (Grêt-Regamey et al. 2017, Li et al. 2020, transport and mobility (Ghent 2018), and tourism (Taff et al. 2019) have a strong effect on ES capacity over time, the workshop results suggested that these changes also influence demand for ES. For instance, limiting air and port (cruise ships) traffic would lower the demand for cultural ES, and thus, the pressure on the current GBI; further, it would reduce air pollution and greenhouse gas emissions, thus reducing the need for urban GBI to provide these regulatory ES.
Furthermore, although the impacts of ES on human health and well-being are at the center of many studies, suggested measures also include improved long-term observation and monitoring of health benefits and the capacity of GBI to fulfill this need. The discussion elicited from the scenarios clearly showed that, on the institutional side, an integration beyond urban greening and planning is needed, including health, tourism, transport and mobility, and education and awareness. These sectors need to be connected to decision-making about land use and quality of life (Webster 2007, Bendt et al. 2013, Tozer et al. 2020). This conclusion points to the role of collaboration in fostering a cross-sector understanding of urban areas as complex social-ecological and adaptive systems, through enhanced crossscale approaches (principles 4 and 5). Thus, the role of institutions in enabling the flow of GBI benefits is not limited to urban planning, land-use policies, and building rights, but extends to other actors and policy sectors that become co-responsible enablers or inhibitors of GBI benefits in the city. In recognizing these linkages and using them to track performance over time, the policy measures and adaptations developed to enhance GBI and ES offer more concrete opportunities to increase the sustainability of the urban system as a whole, pointing toward sustainable and resilient trajectories (Elmqvist et al. 2019). Looking at beneficiaries' perceptions, and taking into account that GBI benefits are co-produced and that potential for realizing ES benefits is closely interrelated and shaped by the cultural and institutional context (Andersson et al. 2019), would improve the understanding of urban ES flow and benefits in the face of diverging ES demands and perceptions (see Riechers et al. 2016). This includes the proposal to develop mechanisms that support value creation and recognition of, for example, public land and the capacity of ES (principle 4).
Greater consideration of ES in other policy fields is needed and demanded. Applying the urban ecosystem services resilience assessment matrix to analyze the effects of policies from other sectors (e.g., housing, transport, health) to the urban GBI could also be explored in this light. The proposed scenarios also triggered reflections on uncertainty in planning while eliciting debate, and not always consensus, with regard to future ES resilience. This uncertainty warrants consideration of alternative and flexible approaches (principle 4; Walker et al. 2001) based on constant monitoring, evaluation, and revision, that are not fully addressed in current policies, but which were raised during the workshop (e.g., inclusion of green indicators into the annual health report).

CONCLUSION
Our study provides a tiered, transdisciplinary approach for engaging urban policy-making to foster adaptive resilience of and through urban ES. The approach consisted of assessing the city's baseline and trajectory for GBI and ES policy, developing possible future scenarios, and co-developing pathways to enhance adaptive capacities for urban policies. Application of the urban ecosystem services resilience assessment matrix for tailoring the ES resilience principles (Biggs et al. 2012) and integrating the three filters approach (Andersson et al. 2019) produced valuable results in the case study application, indicate that specific dimensions of ES capacity and demand are especially vulnerable to change, including microclimate regulation, water balances, mental wellbeing, and social cohesion.
The study revealed for the city of Barcelona an explicit need for fostering systemic, iterative resilience thinking and for considering multilayered processes of change and varied feedback loops. Analysis of current policies and plans revealed the city's current trajectory, with regard to sustainability goals, and highlighted the objectives for action that are needed to support a timely and sustainable urban transformation (Elmqvist et al. 2019). Overall, this research further elucidates the need for strengthening awareness and advancing learning among key https://www.ecologyandsociety.org/vol26/iss4/art38/ urban stakeholders and planners regarding urban socialecological systems as complex adaptive systems. While a shared knowledge base strongly supported the stakeholder engagement process, given the participants' previous awareness and understanding of the topic n considering the diversity of experience and background of workshop participants, involving stakeholders from other policy fields and experts in sectors other than sustainability and resilience with diverse knowledge on GBI and ES topics would be beneficial for further exploring, investigating, and enhancing the role of institutional collaboration while also fostering an understanding of urban areas as complex social-ecological and adaptive systems (principle 4).
The process and application of theory to practice highlights the need for adaptive policies to focus not only on reshaping infrastructures, but to address institutions and perceptions, to be proactive in anticipating future needs and perceptions, and to develop new modes of decision making, co-creation, and comanagement of GBI. Based on adapting and applying the seven principles of resilience to the urban realm and connecting them with the three filters, we argue that adaptive policies (as an institutional tool) should shape current infrastructures with the aims of speaking to and reframing the perceptions of beneficiaries, while also acting preventively to anticipate future needs and perceptions. The use of the urban ecosystem services resilience assessment matrix in policy-making and planning, together with a participatory approach using scenario narratives, can function as a lens for understanding the city's trajectory with regard to sustainability and resilience, identifying leverage points in current policies for unlocking the flows of ES from nature to humans ("resilience of what"), dealing with uncertainties ("resilience to what"), and ensuring inclusivity ("resilience for whom").
Responses to this article can be read online at: https://www.ecologyandsociety.org/issues/responses. php/12535

Data Availability:
The data that support the findings of this study are openly available from Appendices 1 to 6 of the same study. More information can be available on request from the corresponding author. ▪ Some residents are affected more than other; especially vulnerable groups are elderly, children, and pregnant woman. Similarly, heavily build-up areas are affected the most. ▪ An increased importance of friendly and climate-resilient open spaces is highlighted. ▪ Policy interventions are needed in order to preserve and foster the capacity of provisions originating from open spaces, as these are becoming critically scarce Barcelona 2025; Global feedback processes have accelerated the global warming trend to an extent unpredicted by most climate scientists. Barcelona became a critically endangered region, exposing its population to risk and uncertainty. Rapidly changing climate manifests its power through a high variability and frequent occurrence of extreme events.

LITERATURE CITED
A number of torrid days (reaching over 33°C) and the duration of droughts increased eight-fold. Intensive urbanization amplifies the urban heat island effect, making the build-up area hotter by 20°C comparing to the surrounding green areas 5 . The number of deaths counts for 2000 human lives per year due to the heat in Barcelona only 6 . Districts with the lowest coverage of green spaces are hit most heavily, as they become unbearably hot. These areas correspond with the poorer population, whereas elderly, children and pregnant women are the most vulnerable groups. Tropical nights exhausts people as they are not able to have a needed rest after a torrid day. Heat waves cause premature births, physical and mental diseases, but also financial shortages of poorer residents due to intense use of cooling technologies; also, resulting in higher energy use and thus intensifying the climate change.
At the same time, the city experiences severe rainfall shortages and thus water scarcity. Furthermore, fire hazards concern not only the Collserola, but appears in the neighbourhoods of Vallvidrera, Tibidabo i les Planes, Horta, Canyelles, or Torre Baró, where thousands of people reside.
By contrast, rainfall gains on intensity in form of a storm events which frequently flood the city; floods and landslides are threatening residents and causing ample economic damages. A 72% of impermeable surfaces 7 and lack of climate-resilient green spaces in Barcelona are boosting impacts to maximum. Another critical water-related issue will soon emerge on the coastline, where the sea level may rise by additional 1m and start to flood the recreational areas.
Citizens' health and wellbeing started to decrease. A discomfort from heat, droughts, fires, or storm events, has led to a higher demand for friendly and resilient open green spaces. Integrated policy action is required to preserve ecosystem provisions that are necessary for security and wellbeing of citizens.