Explaining the adaptation gap in European cities

doi:10.21203/rs.3.rs-4889351/v1

Download PDF

Brief Communication

Explaining the adaptation gap in European cities

https://doi.org/10.21203/rs.3.rs-4889351/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

An adaptation gap exists when adaptation efforts fail to meet adaptation needs. Conceptualizing and operationalizing this gap is challenging, but should start with, as we argue, addressing ‘internal adaptation policy consistency’. Through a comprehensive European study of local adaptation planning and its progress, we demonstrate how consistency checks applied to adaptation plans can conceptually and methodologically advance understanding of the adaptation (implementation) gap and, most importantly, how to reduce it.

Earth and environmental sciences/Environmental social sciences/Climate-change adaptation

Earth and environmental sciences/Environmental social sciences/Climate-change policy

Europe is warming twice as fast as other continents¹, posing substantial social and environmental risks², particularly for cities where about 75% of European people live³. European cities face an adaptation gap, despite planning for adaptation for decades⁴ and both institutional and financial support. According to UNEP^5:VI, an adaptation gap is “the difference between actually implemented adaptation and a societally set goal, determined largely by preferences related to tolerated climate change impacts and reflecting resource limitations and competing priorities”. Conceptualizing and operationalizing this gap remains challenging, mainly because of difficulties in defining adaptation and measuring its baselines, progress, and goals⁶. Moreover, what is socially acceptable depends on the context and may change with socio-ecological, technological, and political conditions in the future^5,7.

Few empirical studies have assessed the adaptation gap using a clear methodological approach and set of defined proxies. Specifically in urban contexts, studies on single⁸ and large-scale U.S. American cities⁹ have quantified the adaptation gap using climate models and government reports to juxtapose estimated adaptation needs and adaptation progress. These studies rely on the idea that the main gap in adaptation exists between potential adaptation options in planning documents and the actual implementation of adaptation actions on the ground^8,10. One of the most elaborate studies to date involved structured expert judgment to evaluate possible dimensions where adaptation gaps manifest, such as potential and actual levels of risk knowledge, planning, action, capacities, evidence on risk reduction, and long-term strategic pathways. The study was applied across 61 local case studies globally¹¹ and represents a breakthrough in the field of adaptation tracking. However, challenges remain in defining societally accepted goals and impact levels. Moreover, existing research falls short of elucidating strategies for effectively reducing the adaptation (implementation) gap.

In this study, we contribute to these discussions focusing on the “internal adaptation policy consistency” question. Based on a European large-scale study on local adaptation planning and its progress¹², we demonstrate how a series of consistency checks applied to adaptation plans could theoretically and methodologically advance the evaluation of the adaptation (implementation) gap and, most importantly, help reduce it.

‘Consistency’ refers to the utilization of a consistent and operational conceptualization of adaptation’^13:847, which is needed to ensure that any documented difference in adaptation over time and space¹⁴ is not a function of definitional inconsistency. ’Internal consistency’¹⁵ refers to the same level of rigorousness in defining adaptation but within the adaptation plans themselves. We argue that the origins of an adaptation gap go back to the adaptation planning process, identifiable through specific indicators in adaptation documentation.

Plan quality studies often serve as indicators of adaptation progress^12,16–18, utilizing the adaptation management cycle as a theoretical foundation. We contend that only through rigorous¹⁹, credible¹⁷ and frequently updated planning routines genuine adaptation progress can be assessed. This is because shifting baselines, targets, and concurrent societal processes complicate comparisons across different environmental states^6,20. Moreover, we argue that the adaptation management cycle has so far been framed too narrowly: as sequential stages rather than a series of interconnected decisions essential for addressing climate vulnerabilities and enhancing local adaptive capacities. There is substantial theory and evidence showing that linking goals and actions is crucial in policymaking^21,22. While ‘coherency’ is typically discussed regarding alignment among policy domains, governmental levels, or organizations^23–25, there appears to be a significant gap in the analytical understanding of internal consistency within the adaptation planning domain.

We argue that alignment of adaptation planning phases is crucial to successfully plan for and achieve adaptation outcomes¹², as inconsistencies may lead to an adaptation gap. Previous research has highlighted misalignments between adaptation objectives and actions ²⁶. However, other inconsistencies can also indicate an adaptation gap. We identify five key forms of consistency¹²: consistency between risks of hazards and adaptation goals (C1), between sectoral risks and adaptation measures (C2), between risks for vulnerable groups and related measures (C3), between measures for vulnerable groups and related monitoring and evaluation (M&E)(C4), and between risks for vulnerable groups and participation (C5). We use data from a sample of 167 European cities that allow for the analysis of C1 to 5, reported below (Fig. 1).

Across C1 to C5 (Fig. 1, bottom left), European cities show the highest consistency in C2 with 52% of plans fully aligning sectoral risks with adaptation measures. The lowest consistency is observed in C5, where only 1% of plans effectively involve vulnerable groups in plan development. This translates into an adaptation gap of 70%, indicating that more than two-thirds of plans are misaligned, with significant discrepancies particularly in planning for social climate vulnerabilities.

Looking at individual items within C1 to C5, cities exhibit the highest consistency in C2, aligning environmental risks with respective adaptation measures such as improving urban greenery and biodiversity (83% of plans for this item; respective brown bar in Fig. 1) and aligning risks in the water sector with respective actions (80% of plans). Conversely, cities falter in social vulnerability-related items, such as planning for gender- and ethnicity-related risks (rosé bars of C3 in Fig. 1; between 41% and 100% inconsistency), and in including vulnerable groups such as the elderly, children, low-income and ethnic communities in plan development (rosé bars of C5 in Fig. 1; between 92% and 100% inconsistency). Moreover, cities see challenges of including these groups in monitoring and evaluation (M&E) of measures (rosé bars of C4 in Fig. 1; between 81% and 100% inconsistency, respectively), highlighting significant gaps in inclusivity and targeted actions.

The analysis also reveals that many adaptation goals and measures lack preceding risk assessments, (the grey part of bars visible in C1, C2), particularly in sector-specific planning where 49% of decisions are made without risk consideration, compared to 52% that are risk-assessed. Notably, in the environment domain, 92% of plans fail to consider environmental risks before planning measures. This not only illustrates the interdependencies between sectors and urban functions—where the effects of measures may impact different sectors—but also risks leading to maladaptation^27,28, sunk costs, or ineffective adaptation efforts. This widespread discrepancy underscores the need for more rigorous and integrated risk assessment processes in adaptation planning..

Spatial variations of consistency forms unearth regional disparities of a potential adaptation gap across European cities (Fig. 2). Northern European and UK cities show low scores of C1, suggesting high potential adaptation gaps due to misalignments between risk oof hazards and adaptation goals. Conversely, Eastern European cities exhibit lower scores in C3, C4, and C5, which focus on the sensitivities of vulnerable groups, including their risks and impacts, focused measures, monitoring, and participation. Interestingly, Eastern European adaptation plans are more recent, typically post-2018, whereas those in the UK and Northern Europe are older. This suggests that newer plans increasingly neglect the needs of vulnerable groups, revealing that merging different consistency metrics can mask significant regional differences (Fig. 2, bottom left panel).

We acknowledge uncertainties due to the varying thoroughness in documenting the adaptation process, such as risk assessment and participation processes, and in executing outlined measures and actions, monitoring and evaluation. Moreover, we assume that an identified risk indicates a societal goal to reduce it, though the preferred extent of reduction remains unspecified. Future research would benefit from utilizing high-quality adaptation plans with more explicit goals, allowing for a more precise assessment of the quantitative adaptation gap. However, conceptual challenges persist with cascading and complex risks where measures in one sector impact other sectors. These situations often require quantitative modeling of impact networks.. The strength of our methodology, however, lies in the application and quantification of readily accessible, qualitative data.

Our analysis reveals that 70% of adaptation plans in European cities exhibit inconsistencies, minimizing the adaptation potential of those plans and explaining the current adaptation gap. A significant portion of these discrepancies arises from insufficient consideration of social vulnerability, specifically in planning to implement measures for vulnerable groups, and their involvement in plan development and evaluation processes. This oversight raises equity and justice concerns. Many cities across Europe¹² and the US²⁹ continue to struggle with planning for just urban transitions. Our study provides a practical reference point and guidance for addressing these issues. We strongly recommend that decision-makers focus on enhancing internal consistency in future adaptation plans to narrow the existing adaptation gap.

We follow a qualitative research approach relying on data from Reckien et al. (2023), published at https://doi.org/10.17026/dans-xd6-w7pc. Reckien et al. (2023) employed policy/ planning document analysis that allows for interrogation of coherency/ consistency within planning documents.

The city sample

Reckien et al. (2023) covered a representative sample of 327 cities in the former 28 Member States of the European Union (EU-27 plus the United Kingdom) (EU, 2011). For an elaborate description of the data see https://doi.org/10.1038/s42949-023-00085-1. The selected cities are part of the Urban Audit (UA), a Europe-wide initiative for statistical monitoring of major European cities since 2003. The UA cities were selected by the European Commission, Eurostat and the national statistical offices based on the following criteria: (i) approximately 20% of the population is covered in each country; (ii) national capitals and, where possible, regional capitals are included; (iii) large (more than 250,000 people) and medium-sized urban areas (roughly 50,000 and maximum 250,000 population) are to be included; and (iv) urban areas should be geographically dispersed within countries. The UA cities are therefore a balanced and regionally representative sample of cities in Europe. We analyze all UA core cities of the former 28 European countries. Authors have worked in and are familiar with the language and respective urban and climate policies.

The sample of planning document

The policy/ planning documents used by Reckien et al. (2023) are classified as dedicated local climate plans, addressing climate issues as their central topic, as single adaptation plans or joint adaptation-mitigation plans, or broader resilience and sustainability plans (Reckien et al., 2018). They were identified through web search in search engines (Google) and on municipal websites. The plans cover an entire local community/ municipality (not only for parts of the community) and take a comprehensive approach to adaptation/ resilience/ sustainability by focusing on more than just one or two topics or adaptation sectors (that is, we exclude sector-based adaptation plans). Their coding is based on all related and relevant documentation regarding adaptation available, e.g. the adaptation strategy, an action plan, an implementation plan, and/or a monitoring plan even though only the strategy or the action plan might be formally adopted by the local council/ municipality. If multiple subsequent plans were available, the authors analyzed the last updated plan(s). The cut-off date for including adaptation plans/ policies was December 2020 (exact dates visible in the published dataset).

The coding structure

Reckien et al. (2023) coded the adaptation plans in line with the idealized adaptation planning and implementation process (Woodruff and Stults, 2016, EEA) according to the topics of

Fact base: impacts, risks, vulnerabilities

Goals of adaptation

Adaptation measures/ policies/ actions

Implementation tools, processes, timing

Monitoring & Evaluation

Participation in plan development and

general information of the plans and their governance context, e.g., the year of adoption, publication, or approval of the plan; the availability of a national adaptation framework/policy; and the availability of previous plans. After pilot testing, reliability and validity checks, Reckien et al. (2023) employed content analysis (Krippendorf, 2018; Columbia University, 2021) and coding. After coding, the data underwent cleaning and interrater reliability checks.

In line with Reckien et al., 2023, we call the aspects in the list ‘adaptation plan quality principles I) to VI)’. For each of those principles, the author team developed a detailed coding matrix. A detailed overview of the plan topics considered and the possible response options can be found in Table 1.

Their analysis of the adaptation progress of European cities included a conceptualization and calculation of internal adaptation policy consistency.

Here, we argue that the five forms of internal adaptation policy consistency are a perfect indicator for calculating a potential adaptation gap: that means, if adaptation goals are not set according to outlined risks, or if outlined measures are not informed by sectoral and social vulnerabilities, or if M&E is not targeted towards measures focused on those groups most affected, or if participation is not organized along people most in need/ affected by adaptation, we assume the adaptation process and hence its implementation will widen the adaptation gap. Along those lines, respective questions and their matching response options as outlined in Table 1 can be checked for consistency C1 to C5.

Data analysis

The potential adaptation gap was identified by separately revealing inconsistencies from C1 to C5. After thoroughly revising each plan, a binary codification was applied to distinguish the addressed or missing aspects of each principle. “1” values refer to addressed aspects, such as CC impacts, vulnerable sectors, and populations; in addition, the inclusion of vulnerable social groups in M&E and participation phases. “0” values indicate that specific aspects within the form of consistency mentioned are not considered.

For quantitative purposes, we applied pairwise analysis by devising multiple IF functions to distinguish inconsistent and consistent aspects within each principle.

Furthermore, we identified a specific case where the second aspect is present without the previous planning step (i.e., addressed actions related to a previously undefined risk). The theoretical maximum value of potential adaptation gaps was 6, 12, 10, 10, and 10 from C1 to C5, respectively.

Data is available upon request.

We did not compare goals and measures, but risks (hazards for vulnerable groups) and measures. This is because goals are usually so generally formulated^12,26 and so shortly outlined that they hinder a useful comparison to adaptation measures. The risk level was used as a substitute for goals, assuming that an identified risk indicates a societal goal to reduce it, but not how much of a reduction would be societally preferred or tolerated. The question of tolerated impacts, preferences, resource limitations, and abilities remains untouched.

Author Contribution statement

D.R., A.B. and M.A. conceptualized the study. D.R. and A.B. conducted the statistical analyses. D.R., A.B., P.F., and M.S. developed the figures. All authors contributed to writing and editing.

Inclusion & Ethics statement

This research does not involve human subjects but some human subject-related data. To our knowledge, data generation, i.e., coding, data analysis, and data interpretation, do not involve potential sensitivities regarding inclusion and ethics.

Data availability statement

The datasets analyzed during the current study are available at https://doi.org/10.17026/dans-xd6-w7pc.

World Meteorological Organization (WMO). State of the Climate in Europe 2022. (WMO, 2023).
IPCC. Climate Change 2022: Impacts, Adaptation and Vulnerability. (Cambridge University Press, 2022).
Eurostat. Urban-rural Europe. (2024).
European Environment Agency. Urban adaptation in Europe: what works? EEA Report 14/2023. (2024).
UNEP. Adaptation Gap Report 2023: Underfinanced. Underprepared. Inadequate investment and planning on climate adaptation leaves world exposed. (United Nations Environment Programme, 2023). doi:10.59117/20.500.11822/43796.
Garschagen, M. et al. Cross-Chapter Box PROGRESS. Approaches and Challenges to Assess Adaptation Progress at the Global Level in Decision-Making Options for Managing Risk. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate. in Climate Change 2022 – Impacts, Adaptation and Vulnerability 2539–2654 (Cambridge University Press, 2022). doi:10.1017/9781009325844.026.
Bulkeley, H. The condition of urban climate experimentation. Sustain. Sci. Pract. Policy 19, (2023).
Chen, C., Doherty, M., Coffee, J., Wong, T. & Hellmann, J. Measuring the adaptation gap: A framework for evaluating climate hazards and opportunities in urban areas. Environ. Sci. Policy 66, 403–419 (2016).
Lu, S. & Nakhmurina, A. Financial constraints and short-term planning are linked to flood risk adaptation gaps in US cities. Commun. Earth Environ. 5, 43 (2024).
Arteaga, E., Nalau, J., Biesbroek, R. & Howes, M. Unpacking the theory-practice gap in climate adaptation. Clim. Risk Manag. 42, 100567 (2023).
Magnan, A. K. et al. Status of global coastal adaptation. Nat. Clim. Chang. 13, 1213–1221 (2023).
Reckien, D. et al. Quality of urban climate adaptation plans over time. npj Urban Sustain. 3, 4–6 (2023).
Ford, J. D. & Berrang-Ford, L. The 4Cs of adaptation tracking: consistency, comparability, comprehensiveness, coherency. Mitig. Adapt. Strateg. Glob. Chang. 21, 839–859 (2016).
Gibbs, M. T. Consistency in coastal climate adaption planning in Australia and the importance of understanding local political barriers to implementation. Ocean Coast. Manag. 173, 131–138 (2019).
Nowak, A. C. et al. Opportunities to strengthen Africa’s efforts to track national-level climate adaptation. Nat. Clim. Chang. (2024) doi:10.1038/s41558-024-02054-7.
Meerow, S. & Woodruff, S. C. Seven Principles of Strong Climate Change Planning. J. Am. Plan. Assoc. 86, 39–46 (2020).
Olazabal, M., Galarraga, I., Ford, J., Sainz De Murieta, E. & Lesnikowski, A. Are local climate adaptation policies credible? A conceptual and operational assessment framework. Int. J. Urban Sustain. Dev. 11, 277–296 (2019).
Woodruff, S. C. & Stults, M. Numerous strategies but limited implementation guidance in US local adaptation plans. Nat. Clim. Chang. 2016 68 6, 796–802 (2016).
Preston, B. L., Westaway, R. M. & Yuen, E. J. Climate adaptation planning in practice: An evaluation of adaptation plans from three developed nations. Mitig. Adapt. Strateg. Glob. Chang. 16, 407–438 (2011).
Dupuis, J. & Biesbroek, R. Comparing apples and oranges: The dependent variable problem in comparing and evaluating climate change adaptation policies. Glob. Environ. Chang. 23, 1476–1487 (2013).
Knill, C. & Tosun, J. Policy Making. in Comparative Politics (6th edition) (ed. Caramani, D.) (Oxford, OUP, 2023).
Walker, W. E. Policy Analysis: A Systematic Approach to Supporting Policymaking in the Public Sector. Decis. Anal 9, 11–27 (2000).
Smith, G., LeTissier, M., O’Hagan, A. M. & Farrell, E. J. Policy Coherence for Climate Change Adaptation at the Land-Sea Interface in Ireland. Plan. Pract. Res. 37, 173–188 (2022).
Kivimaa, P. et al. Evaluating policy coherence and integration for adaptation: the case of EU policies and Arctic cross-border climate change impacts. Clim. Policy (2024) doi:10.1080/14693062.2024.2337168.
Roy, M. et al. Quality of Cities’ Networks of Plans and Prospects for Flood Resilience. J. Plan. Educ. Res. (2024) doi:10.1177/0739456X241236486/ASSET/IMAGES/LARGE/10.1177_0739456X241236486-FIG2.JPEG.
Goonesekera, S. M. & Olazabal, M. Climate adaptation indicators and metrics: State of local policy practice. Ecol. Indic. 145, 109657 (2022).
Juhola, S. & Käyhkö, J. Maladaptation as a concept and a metric in national adaptation policy- Should we, would we, could we? PLOS Clim. 2, e0000213 (2023).
Reckien, D. et al. Navigating the continuum between adaptation and maladaptation. Nat. Clim. Chang. 13, 907–918 (2023).
Caggiano, H., Kocakuşak, D., Kumar, P. & Tier, M. O. U.S. cities’ integration and evaluation of equity considerations into climate action plans. npj Urban Sustain. 2023 31 3, 1–9 (2023).

Table 1

Plan topics with questions and respective response options, along plan quality principles, as used in this study. Data available upon request.
Plan quality principles used in this study	Plan topics and related questions that are used as indicators of plan quality in this study	Response options (in abbreviated form)
I. Fact base - Climate change impact, risk and vulnerability assessment	General CC impacts/ risks: Does the plan include a discussion of the global or national impacts/ risks of CC?	Yes - No
	Specific CC impacts/ risks: Does the plan include a discussion of the specific impacts of CC to the jurisdiction and, if yes, related to which timescale?	None - Past – Present – Future
	Specific CC impacts/ risks: Which impacts, if any, are identified in the jurisdiction?	None – Temperature variation – Precipitation variation – Storms/ wind variation – Coastal flooding – Inland flooding – Drought – Landslides - Other
	Equity/ justice of impact/ vulnerability assessment: Does the plan identify certain population groups that are/ will be particularly impacted by CC and, if yes, which ones?	None – Women – Infants – Elderly – Poor – Sick – Immobile – Lone parents – Ethnic communities – Migrants – “Vulnerable persons” - Other
	Does the plan identify certain sectors, if any, that are/ will be particularly impacted by CC?	None – Industries general – Buildings – Transport – Energy – Water – Waste – Agriculture/ Forestry – Environment/ Greenery – Health – Social institutions – Civil protection – Tourism – Financial institutions
II. Adaptation goals	Specific adaptation goals: Does the plan identify at least one concrete adaptation goal to a particular CC impact/ risk and, if yes, to which impact/ risk?	None – Temperature variation – Precipitation variation – Storms/ wind variation – Coastal flooding – Inland flooding – Drought – Landslides - Other
II. Adaptation goals	Quantitative adaptation goals: Does the plan identify at least one quantitative adaptation goal? If yes, specify	Open field
III. Adaptation measures	Buildings (and related land use planning): Does the plan include at least one measures in the building sector and, if yes, which ones?	None – Building codes – Insulation – Compact development – Low-density development – Mixed use – Brownfield development - Other
	Transport: Does the plan include at least one measure on transportation and, if yes, which ones?	None - Public transport – Bicycle-oriented development – Pedestrian-oriented development - Other
	Energy: Does the plan include at least one measure on energy and, if any, which ones?	None – Solar roofs – Hydropower - Other
	Water: Does the plan include at least one measure related to water management and, if yes, which one?	None – Water infrastructure – Water conservation - Other
	Waste: Does the plan include at least one measure on waste management?	Yes - No
	4.6 Agriculture and forestry: Does the plan include at least one measure on food security and agriculture and, if any, which ones?	None – Land conservation – Support of farmers – Support for organic food - Other
	Environment, greenery, biodiversity: Does the plan include at least one measure on resource mgt., green infrastructure, ecosystem services, if any, which ones?	None – Green spaces – Blue/ Green spaces – Blue spaces – Brown spaces - Other
	Health: Does the plan include at least one measure on the social and environmental determinants of health or social care provision systems and, if any, which ones?	None – Health determinants – Care systems - Other
	Social and educational institutions and services: Does the plan include at least one measure on social institutions and, if yes, which ones?	None – Education facilities – Sport facilities – Care facilities - Other
	Civil protection and emergency: Does the plan refer to the development or use of disaster response systems?	Yes - No
	Tourism: Does the plan include at least one measure related to tourism?	Yes - No
	Industries/ economies/ businesses/ financial institutions: Does the plan include at least one measure for particular industries?	Yes - No
	Equity/ justice: Are specific population groups mentioned in relation to any of these measures above?	None – Women – Infants – Elderly – Poor – Sick – Immobile – Lone parents – Ethnic communities – Migrants – “Vulnerable persons” - Other
IV. Implementation tools, process & progress	Priority: Does the plan prioritize certain actions for implementation and, if yes, how?	None – Sequence, without reasoning – General importance, without reasoning – Highest/ Largest anticipated impacts – Best-practice options – Economic feasibility – Synergies/ Conflicts – Easy/ Medium/ Difficult implementation – Effectiveness for impact/ vulnerability reduction - No regret options – Low regret options, i.e. “low hanging fruits”- Win-win options – Flexible, adaptive options – Multi-benefit options (including, CO2 reduction)
	Responsibility: Does the plan identify a responsible party for at least one specific adaptation measure?	Yes - No
	Timeline: Does the plan identify timelines for implementation?	Yes - No
	Budget: Does the plan include considerations of human and financial resources required and/or a budget for implementation?	Yes - No
V. Monitoring & evaluation	M&E: Does the plan include a section or separate plan that describes the envisaged M&E activities?	Yes - No
	Responsibility: Does the plan identify responsible parties for M&E and, if yes, which ones?	None – Municipality: Same department like plan development – Municipality: Different department – University/ Research centre/ Scientific community – External consultants/ Companies/ Businesses/ Industries – NGOs - Other
	Objectives: Does the plan include M&E objectives?	Yes - No
	Indicators: Does the plan include M&E indicators?	Yes - No
	Timeline: Does the plan include a timetable for M&E?	Yes - No
	Equity/ justice: Does the M&E process consider assessing the outcome of adaptation on population groups and, if yes, which ones?	None – Women – Infants – Elderly – Poor – Sick – Immobile – Lone parents – Ethnic communities – Migrants – “Vulnerable persons” - Other
VI. Participation	Participation: Does the plan identify the organizations and stakeholders involved in the plan making process and, if yes, which ones?	None – Municipality: Different department – University/ Research centre/ Scientific community – The public – Community and neighbourhood groups – External consultants/ Companies/ Businesses/ Industries – NGOs – Other
VI. Participation	Participation, equity/ justice: Is there mentioning of particular population groups being involved/ heard during the plan making process, if any?	None – Women – Infants – Elderly – Poor – Sick – Immobile – Lone parents – Ethnic communities – Migrants – “Vulnerable persons” - Other

There is NO Competing Interest.

SupplementaryInformation2.docx
Supplementary Information

Download PDF

Version 1

posted

You are reading this latest preprint version

Explaining the adaptation gap in European cities

Status:

Version 1

Abstract

Figures

Main Text

Online Methods

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1