Enabling spaces for bridging scales: scanning solutions for interdisciplinary human-environment research

To tackle the twenty-first-century challenges for sustainability, a deeper understanding of their complexity is needed. Hence, interdisciplinary human–environment research integrating knowledge, perspectives, and solutions across scales is crucial. Yet, questions of ‘scale’ and ‘scaling’ continue to challenge human–environment research. Despite substantial scholarly attention to scales, no unified definitions and understanding exist, causing confusion among researchers. In this paper, we aim to provide clarity to the challenges and potential solutions to bridging scales in interdisciplinary human–environment research as well as identify leverage points and arenas of change that would enable it. We used a three-step methodology to (a) identify the challenges and solutions based on a survey of 82 interdisciplinary researchers, (b) prioritise the challenges and solutions based on expert elicitation, and (c) organise the solutions according to their potential for leveraging change at three system arenas: individual researchers, institutions and organisations, and the wider academic structures. The main challenges ranged from issues related to differences in research approaches, lack of resources for deeper collaborations, or conflicting perspectives on problem and scale definitions, to miscommunication. The specific solutions linked to these challenges were predominantly related to the potential of improving the research process through open communication, data sharing, or new techniques for upscaling or downscaling results. We argue that bridging scales requires open and humble conversations across disciplines on scale understandings as well as improved collaboration through data sharing, method development, result dissemination, and contribution to theory. Facilitating such collaboration requires enabling spaces of interaction across three systemic arenas: the individual researcher’s personal commitment to open communication and reflection, the research institution’s capacity to enable interdisciplinary spaces, and the wider academic system valuing and supporting interdisciplinary and cross-scale initiatives.


Introduction
It is widely acknowledged that addressing the inherent complexity of today's sustainability challenges including climate change, biodiversity loss, food insecurity, and social and economic inequality requires cross-scale and interdisciplinary collaboration (Díaz-Reviriego et al. 2019;Meyfroidt et al. 2022;Singh et al. 2012). Indeed, the diverse fields of study that can be subsumed under the heading of "human-environment research" are characterised by efforts to integrate insights from the local to the global (Pereira et al. 2020;Reid 2006;Verburg et al. 2013), as well as the natural and social sciences (Fox et al. 2006; Moran and Lopez 2016;Schneidewind et al. 2016). Yet, despite substantial scholarly attention to when and how integration and collaboration across disciplinary boundaries and knowledge domains can be fruitful (Barry and Born 2013;Choi and Richards 2017;Darbellay 2015;Manson 2008;Miller et al. 2008;Reid et al. 2006), achieving cross-scale and interdisciplinary collaboration in practice continues to challenge human-environment researchers on a day-to-day basis (Cairns et al. 2020;Freeth and Caniglia 2020;Rocha et al. 2020;Villeneuve et al. 2020).
One of the particularly challenging issues related to such effort is different conceptualisations of 'scale', scale terminology, and approaches to the processes of 'scaling' data, results, and knowledge (Cash et al. 2006;Gibson et al. 2000;Manson 2008;Pulver et al. 2018). In much of the human-environment research, 'scale' is commonly used to refer to the variety of "spatial, temporal, quantitative, or analytical dimensions used by scientists to measure and study objects and processes" (Gibson et al. 2000, p. 219) with the term 'level' used to denote "the units of analysis that are located at different positions on a scale" (Cash et al. 2006, p. 2). Yet, the terminology used to describe scales differs among disciplines. In ecology and to some degree in physical geography, scale is often defined in terms of grain and extent, denoting the resolution of the measurements or observation in data and analysis, and the spatial or temporal magnitude over which observations are made (e.g. study area or duration of the study) (Sayre 2005;Hay 2005;Turner et al. 1989). Scale is also used to refer to measurement tools and instruments in experimental and quantitative research (Brunsdon 2018), while scale is understood and heavily debated as one dimension of complex socio-spatial relations in human geography and socio-spatial theory (Brenner 2001;Jessop et al. 2008;Marston 2000). Moreover, while the denominators 'small-scale' and 'large-scale', for example, are commonly used to refer to small or large phenomena with regard to scales of space, time, or quantity, cartographers and others concerned with mapping environmental change use the terms in the sense of map scales, i.e. smallscale refers to maps covering large areas (e.g. continents), while large-scale maps represent smaller areas (e.g. a city district) (Goodchild 2011). Such terminology differences give rise to confusion and risk hampering interdisciplinary collaboration in practice.

The scale challenge for interdisciplinary humanenvironment research
While it is beyond the scope of this introduction to present an in-depth review of the diverse scholarly debates on scale in different disciplines, we see at least three barriers for interdisciplinary human-environment research seeking to work across scales. First, many human-environment interactions are characterised by scale mismatches between the social and ecological processes embedded in them (Cumming et al. 2006). The organisation of human societies in communities, cities, or countries rarely matches the extent of the ecosystems or biophysical processes they act in. For interdisciplinary research, disciplinary conventions for focus of study, conceptual backgrounds, and data sources or availability, challenge common research designs (Dirnböck et al. 2013). Political scientists and economists interested in human-environment issues, for example, often study human decision-making and behavioural change at various levels of decision-making and may use surveys, official statistics, or census data related to administrative boundaries that predefine the scale of study (Moss and Newig 2010;Veldkamp et al. 2011). Ecologists and physical geographers interested in the patterns, processes, and spatial extent of ecological or landscape change, in turn, may for example rely on observational data from field surveys that determine the grain and extent of their studies to a watershed or biomes (Semper-Pascual et al. 2020), while remote sensing analyses may be framed by the availability of satellite images and their resolution (Moon et al. 2021). Combining the results from such diverse research approaches in cross-scale analysis is thus often challenging.
Second, the process of scaling results and knowledge differs between research domains. Scaling overall refers to the translation of information and knowledge from one scale to another, including upscaling and downscaling results or rescaling governance models (Hay 2005;Newig and Moss 2017). This is both an empirical and analytical challenge, since physical and ecological processes scale differently than social processes (Niewöhner et al. 2016). For example, the ecological properties of a tree, a coppice, and a forest diverge and scale in completely different ways than do the social properties of an individual, a family, and a society. Upscaling or downscaling data or results, moving from the specific to general understanding, establishing causal relations, or making identified solutions applicable at different scales thus require substantially different methodological and theoretical approaches depending on the object of the study.
Finally, at a more fundamental level, epistemological approaches to the concept of 'scale' differ among disciplines from realist to constructionist perspectives (Manson 2008;Marston 2000;Moore 2008;Zhang et al. 2014). Realist scale perspectives rely on the premise that scales and scale levels exist independently of the object or process being studied and of the observer (Sayre 2005). Human-environment interactions are viewed as scale dependent in the sense that certain phenomena and processes occur naturally at certain scales, such as habitats for different species. Analytically, the challenge is to identify the best apparent fit between the scale of analysis and the specific change or process of interest (Manson 2008). Realist accounts of scale are generally used in the natural sciences and are complemented by hierarchy theory that contributes with a distinction between absolute scales with independent levels that can be objectively measured, and relative scales with interdependent levels related to and defined by the object or process under study (Gibson et al. 2000;Manson 2008). In contrast, constructionist perspectives on scales are based not only on the proposition that knowledge about the world is socially constructed, but also that scales are actively created through the research and through political processes that over time manifest in material reality as well (Manson 2008;Marston 2000). Scale terminology and metaphors such as local, regional, national, and global are thus not neutral or objective, but embedded in discursive and political power relations. It is therefore crucial to examine how such metaphors are constructed for a particular analytical or political purpose, rather than understanding them as given a priori (Beck et al. 2017;Howitt 1993Howitt , 1998. Although constructionist notions of knowledge production in human-environment research acknowledge that a material reality exists, this reality can never be divorced from the concepts and categories used to describe it, including analytical scales (Jones 2002;Niewöhner et al. 2016). Human geographers, anthropologists, political ecologists, and political scientists, for example, have used constructionist perspectives on scale to show how particular scales can be politically and linguistically constructed and manipulated with material impacts in human-environment systems (Marston et al. 2005;Neumann 2009;Newig and Moss 2017;Swyngedouw 2004).

Bridging scales
The discussions about these fundamental epistemological differences around scales, and their manifestations in different terminology, scaling challenges and scale mismatches are not new (Ahlborg and Nightingale 2012;Reid 2006). Efforts to bridge both spatial scales and knowledge domains were already at the centre of the Millennium Ecosystem Assessment in the early 2000s (Hay 2005;Reid et al. 2006). Yet, they continue to challenge researchers seeking interdisciplinary collaboration on complex human-environment relations. Recently, debates in the context of the Intergovernmental Science-Policy Platform for Biodiversity and Ecosystem services (IPBES) related to Nature's Contribution to People are advancing the scale bridging efforts by presenting novel ways of weaving together generalised and context-specific perspectives at multiple scales (Díaz et al. 2015;Hill et al. 2021).
In this paper, we join these debates with the aim of examining the day-to-day problems and tensions, both theoretical and practical, that human-environment researchers face when dealing with cross-scale issues, as well as to scan solutions that enable spaces of collaboration for 'bridging scales'. Our aim is thus not to 'settle' any (inter)disciplinary discussions on the definition of scales and scaling-which is neither possible, nor necessarily desirable-but rather to use the notion of 'bridging scales' as a starting point for creating a better understanding of plural perspectives on scale in order to improve interdisciplinary collaboration in terms of theory, data, methods, and findings. A central argument is thus that we need to identify leverage points for change (Abson et al. 2017;Dorninger et al. 2020;Meadows 1999) in the current academic system to foster spaces for exchange and learning that enable a diverse scale understanding.
We take our cue for this collaborative research from the challenges we have encountered as a group of early-career researchers from diverse disciplinary backgrounds from human and physical geography, through hydrogeology, forestry and conservation science, and remote sensing, to environmental policy and institutional economics. Through our collaborative work at the Integrative Research Institute for Transformation of Human-Environment Research (IRI THESys) at the 1 3 Humboldt-Universität zu Berlin, we encountered several points of tension related to our different understandings of scale as a concept and a research frame. In this paper, we use these tensions and the perspectives presented above as a starting point to pursue three main objectives. First, we identify the main challenges for bridging scales among human-environment researchers; second, we identify a set of solutions for those challenges that can advance interdisciplinary discussions and practical collaboration in human-environment research. Finally, we analyse the solutions according to their leverage potential for change across three systemic arenas in the academic system, namely individual researchers, research institutions and organisations, and the wider academic structures.

Methods
To address our objectives, we applied a three-step methodology (Fig. 1a). We used an extended version of the classical solution-scanning exercise to first identify the challenges experienced by human-environment researchers when bridging scales, and second to determine the most suitable solutions, before asking experts to prioritise these solutions. Solution scanning is a stepwise methodology used to identify a list of actions, interventions, or approaches that respond to a challenge, and that can be used to set research agendas, or identify potential policy interventions in decision-making processes (Dicks et al. 2017;Sutherland et al. 2014). Solution scanning proceeds by first defining an objective or a challenge. The objective can be deduced from knowledge gaps in the literature, or from societal needs and concerns (Pullin et al. 2013). Then a group of experts is asked to use their experiences to identify actions that can leverage change towards the stated goal. In the last step, the proposed solutions are listed and distributed to the experts to be assessed and prioritised according to given criteria (Hernández- Morcillo et al. 2018;Plieninger et al. 2018).
Following the solution-scanning exercise, we used the notion of leverage points to analyse the potential of the identified solutions to induce changes in the academic system that could facilitate bridging scales in interdisciplinary human-environment research. The leverage point notion is based on a system perspective, where interventions and solutions are considered based on their capacity to induce shallow or deep changes (Abson et al. 2017;Dorninger et al. 2020;Meadows 1999). In her seminal paper on leverage points, Meadows (1999) identified twelve types of system interventions that were later synthesised by Abson et al. (2017) into four types of system characteristics organised from the shallowest to the deepest: parameters, feedbacks, design, and intent. Parameters are modifiable, mechanistic characteristics or physical elements of a system such as incentives, standards, or factors typically targeted by decision-makers and individuals. Feedbacks are the interactions between elements within the system that drive internal dynamics or provide information regarding desired outcomes. Design characteristics relate to the social structures of information flows, rules, power, and selforganisation that manage parameters and feedback. Finally, intent characteristics relate to the norms, values, and goals embodied within the system of interest and the underpinning paradigms from which they arise (Abson et al. 2017).

Exploratory survey design
To gain a broad understanding of the challenges interdisciplinary human-environment researchers are facing in relation to scales and scaling, as well as to elicit as many ideas for solutions as possible, we conducted a survey among the participants of the KOSMOS Conference "Navigating the Sustainability Transformation in the 21st Century" organised by the Humboldt-Universität zu Berlin in August 2019. 1 The KOSMOS Conference was attended by 316 people, primarily researchers and students, as well as a few practitioners, from a broad range of fields to discuss today's great challenges for sustainability, and hence presented a good setting to carry out the survey.
The survey was designed in four parts: (1) understanding of the term 'scale' and relevant 'scales' in the respondents' research; (2) listing the three main challenges in relation to bridging scales; (3) identification of the three most effective solutions related to the named challenges; and (4) characterising respondents' profiles including education, current research topic, and biographical data.
A total of 82 people participated in the survey (25.9% of all participants) with an almost equal gender distribution (50% female, 48% male, and 2% not specified). The vast majority were academic professionals and students (90% of the respondents), and most of the respondents came from or had their primary employment in Germany (74%). The respondents' reported disciplines were mainly sustainability (27%), natural (23%), and social sciences (12%). A quarter of respondents declared that they work in interdisciplinary settings. When considering their current field of research, 1 3 many respondents with a background in the socio-economic, agriculture, and forestry sciences indicated that they had moved from a disciplinary background into more interdisciplinary work.

Coding understandings and types of scales used by the respondents
To analyse the survey responses regarding 'understanding of scales', we coded the full answers from the survey (n = 82) together as a team. In a first round, answers that were conceptually similar were grouped together under a conceptual category. Some answers included more than one component and were coded under more than one category, which resulted in a total of 137 entries. In a second round of coding, the categories with less than two entries were re-organised and merged into other categories. We then consolidated the categories in a consensual process (Fig. 1a, see also Table 1 in the Supplementary Material).
To analyse the 'types of scales used by the respondents', we counted the frequency of identical or similar answers and grouped these into categories. Also here, several answers were placed into more than one category resulting in a total of 321 entries. We then consolidated the categories (Fig. 1a, see also Table 2 in the Supplementary Material).

Coding of challenges and solutions
Before coding the challenges and solutions, we disaggregated survey responses containing multiple components into individual statements (Fig. 1b, step 1). This resulted in 286 individual challenges and 288 individual solutions. We then coded the challenges and solutions manually in several rounds of coding. In the first round, we individually categorised the statements and proposed overarching categories (step 2). Subsequently, we collaboratively merged similar categories identified by different coders agreeing on a common set of categories (step 3). In a second round of coding, we discussed the content of the emerging categories, reassigned and recoded some statements, and grouped thematically similar categories under the same overarching heading to consolidate challenges and solutions into main categories and subcategories (step 4). Two of the main solution categories contained statements that were aligned in content to the extent that subcategories were not necessary. At this stage, we also discarded several statements containing vague or unclear answers, such as incomplete sentences or one-word answers. Finally, in a third round of coding, we split into two groups to check the consistency of the statements coded in each category, and to agree on possible re-categorisations and new categories (step 5).

Expert identification
The second step of our study aimed at connecting and prioritising the key challenges with suitable solutions and involved an online expert elicitation process. We approached experts of different disciplinary backgrounds but working in the field of human-environment research and with expertise in questions of scale and scaling. We identified experts through (1) suggestions made by survey participants; (2) established scholars identified through the literature; and (3) the co-authors' scholarly network. In addition to these criteria, we considered geographical representation, gender balance, and a diversity of disciplinary backgrounds from across the social and natural sciences, as well as interdisciplinary fields such as sustainability science as key features for finding potential experts. Our final list included 59 scholars (39 male, and 20 female), who were approached via email.
Sixteen experts participated in the elicitation process (response rate of 27%). They were four women, eleven men, and one undisclosed, and the majority were from Europe with a few from South and North America, as well as Africa. The experts had a wide range of disciplinary backgrounds within the social and natural sciences, as well as some with interdisciplinary profiles. Several of the experts indicated that they had made a move from more disciplinary fields of research, e.g. mathematics, geography, and economics, during their PhD degree to more interdisciplinary fields of research, including environmental resource management, land system science, and sustainability science (see Table 3 in the Supplementary Material).

Questionnaire design
We used an online survey tool with a questionnaire divided into three sections for the expert elicitation process. First, we asked the experts to choose the three challenges deemed most pressing and to elaborate on the reasons for their choice. For the sake of clarity, we presented the experts with the main challenges categories (n = 8) instead of the subcategories (n = 31). Second, we asked the experts to choose up to three solutions for each of the challenges and to explain their selection. Here, we presented the experts with the full list of solution subcategories (n = 23) to provide a full range of options and allowed the experts to provide suggestions for implementing the solutions. We presented both challenges and solutions as full sentence statements and provided lists elaborating and explaining the content of each category.
Then, we asked which scales they primarily used in their work and finally we asked for profile information.

Coding process and analysis of expert answers
We analysed the experts' answers related to their definition of scales using the coding scheme developed based on the survey responses. While all the experts' answers fell within the identified main scale categories, they also brought attention to several new types of scales, which we coded into two new subcategories.
For analysing the challenges and solutions prioritised by the experts, we first ranked the challenges in terms of number of experts prioritising them and second grouped the selected solutions for each of the challenges. We qualitatively analysed and compared explanations for the choice of challenges and the elaborations related to solutions.

Third step: organising the solutions according to their leverage potential
To cluster the different solutions according to their potential for system-wide change and sustainability transformation, we analysed the prioritised solutions resulting from the expert elicitation using the notion of leverage point to sustainability transformation (Abson et al. 2017;Dorninger et al. 2020;Meadows 1999). For that purpose, we first defined our system of interest as the academic system concerned widely with human-environment interactions. This comprises the different disciplines, research organisations, researchers, communication structures, and funding schemes. Then, we divided the author team into two groups to discuss each solution and assign them into one of the four types of system characteristics: intent, design, feedbacks, and parameters. We then compared both groups' outcomes and reached a final classification by consensus. Within each type of system characteristics, we sorted the solutions according to their perceived relations to the previous and next type.

Scale definitions
The survey respondents' understandings of scale showed the breadth of definitions attached to this concept in human-environment research and as represented by their diverse backgrounds. Several of the respondents noted how their disciplinary training influenced their understanding of scale, while others pointed out how the multiple definitions and uses of scales challenge interdisciplinary research at a fundamental level. One respondent pinpointed this diversity by stating that it is "hard to define the 'term', I guess it depends on one's area and how concrete you want to be. e.g., it can be about relatedness, e.g., of human scale and the universe, more philosophical. Or simply a level, a size of something as in maps" (Survey-ID66).
In the coded responses, the majority referred to specific spatial associations like "local" or "continental" (n = 42) and descriptive terms like "level" and "range" (n = 31), while fewer referred to how 'scale' can be defined by its characteristics depending on the context it is applied in (e.g. referring to size, range or scope) or the relations it is used to describe (e.g. between time and space or material-immaterial world) (n = 15). In addition, a large number of the coded responses linked scale to the research process as a particular perspective or tool of measurement (n = 22), while others more closely related the term to the process of establishing units of analysis (n = 8) and boundaries for the scope of research (n = 7). Finally, a few responses emphasised how we need to engage with scale through more critical perspectives related to power, conflict, and diversity (n = 4). The diversity of responses related to the understanding of scale was mirrored in multiple types of scales used by survey respondents in their own research (Table 1). Spatial scales and levels, e.g. local or global, were by far the most widely used type of scales among our respondents, followed by spatial-political scales, political/administrative scales, and temporal scales. In addition, many respondents defined the type of scale they work on in thematic terms such as "human scale vs. the economy", "global scale of SDG implementation" or "scale of interlinks between consumers and producers" (referred to as "thematic scales" in Table 1).
The experts' use of scale was also very diverse, and largely mirrored that of the survey respondents. The majority of experts defined their work in relation to spatial scales such as local, regional, or global, and several worked on spatial political scales or political/administrative/institutional scales. Moreover, some of the experts indicated that their work mainly related to change and cross-scale interactions including upscaling, downscaling, and local-global interaction. In addition to the categories used by the survey respondents, two of the experts mentioned additional types of scales related to "data" and "social-ecological systems".

Identification of challenges
We identified 31 specific challenges for bridging scales in human-environment research grouped into eight main categories (hereafter, 'main challenges'), of which differences in research approaches, the process of scaling, and communication had the highest frequency in the survey responses (Table 2). Less prevalent, although still noticeable, were challenges related to structural problems within the academic system, differences in problem definition, and problems related to the personalities and profiles of the researchers. Finally, some responses were concerned with issues related to fundamental conflicting perspectives among social and natural researchers, and transdisciplinary challenges seeking to bring knowledge from research into action.

Prioritisation of challenges
The eight main challenges were prioritised between nine and three times by the sixteen experts (Table 2). Also, among the experts, the most prominent challenge was differences in research approaches prioritised by nine of the experts. Here, the experts highlighted how research approaches in different fields often are designed to target different scales making collaboration challenging. Several experts also pointed out that different research approaches make it difficult to establish causality across scales, to balance the level of detail and depth of analysis, as well as generalising results through upscaling or downscaling.
The second most prominent challenge relates to structural challenges embedded in the academic system. The experts pointed to problems in the current funding structures noting for example that limitations on funding and time make researchers focus on a single case for the sake of feasibility. Moreover, it was emphasised that the challenges of bridging scales are closely related to and embedded in the barriers to pursue interdisciplinary research and interdisciplinary research careers in the education system, evaluation structures, and scientific journals.
On a shared third place, three of the challenges were prioritised by six experts. The first was fundamental differences in problem definition, i.e. in the way human-environment interactions are understood, conceptualised, and approached in different fields. Here, it was noted that the inherent complexity of human-environment systems such as scale mismatches between social and ecological processes is a fundamental challenge, complicating the establishment of appropriate system boundaries and units of analysis both in spatial and temporal terms. The other challenge was conflicting perspectives among social and natural scientists. One expert elaborated stating that "conflicts start with the different meanings, definitions, and delineation of 'space' in cultural and physical geography and end in a growing lack of understanding between the social and natural sciences" (Expert-ID11). The challenge is then to negotiate different understandings and find ways of compromising and balancing disciplinary strengths and interdisciplinary collaboration. Lastly, six experts pointed to transdisciplinary challenges related to bringing knowledge from research to action. This was particularly related to scaling up actionable knowledge and bridging knowledge on problems and solutions between actors working at different levels and scales. One of the experts, for example, argued that the problem "goes beyond accessibility of specialised knowledge: it deals with the question of how to generate common understandings among different levels of action" (Expert-ID19). These fundamental transdisciplinary issues are related to the inclusion of different perspectives in problem definition, research agendas, and research finance.
The final three challenges were process of scaling, communication, and the researcher's profile prioritised by five, four, and three experts, respectively. One of the issues highlighted concerning the process of scaling was the clash Table 2 The main categories and subcategories of challenges for bridging scales in human-environment research identified through the survey (N = 264), and the frequencies of coded responses The challenges are ordered according to the experts' prioritisation of challenges (see full explanation of the categories in Table 4

Identification of solutions
From the survey, we identified 23 solutions under seven main categories (Table 3). By far, the survey respondents most frequently referred to solutions related to improving the research process with more than 100 individually coded suggestions. In addition, responses focused Table 3 The main solution categories and subcategories identified through the survey (N = 277), and the frequencies of coded responses (%) for bridging scales in human-environment research (see full explanation of the categories in Table 5 Table 3, but have been shortened for readability on solutions related to improving the research system through e.g. fostering better environments for collaborative research groups and networks, and developing longterm research projects that can enable the interdisciplinary exchange. In addition, suggestions for working towards better communication within and beyond academia featured frequently among the responses, as did solutions related to improving the researchers' profile towards more open-mindedness, humility, and reflexivity among researchers working on interdisciplinary topics. The remaining solutions were grouped around topics such as improving interdisciplinary education and training that foster thinking across scales and increasing funding for inter-and transdisciplinary research.

Linking challenges with solutions and prioritisation of solutions
The results of the expert elicitation showed that the experts prioritised a multiplicity of solutions for the challenges related to differences in research approaches, conflicting perspectives, and from research to action, as well as to the researchers' profiles (Fig. 2). For other challenges, i.e. the process of scaling and communication challenges, the prioritised solutions were focused on a few targeted areas.
Looking across the prioritised solutions, suggestions for improving education that fosters thinking across scales (B1) and providing students with interdisciplinary competences (B2) feature prominently for several of the challenges, mainly for tackling conflicting perspectives and differences in research approaches related to scales. One expert emphasised that "training and education should prepare future researchers to arrive from a defined problem [to] a scalecrossing understanding" and continued by noting that "to be able to integrate data from various perspectives, requires a profound knowledge about scale issues, awareness raising of the scale issues and careful consideration in the design of the research process" (Expert-ID9). For five of the challenges, the experts highlighted the need for institutionalisation of interdisciplinary study programmes that consider diverse scale perspectives (B3). In contrast, one expert noted that while researchers trained interdisciplinarily are great at collaborating across knowledge domains, they often lack the in-depth theoretical and methodological knowledge that is associated with disciplinary training. To abate this, the expert proposed a stronger focus on fostering enlightened interdisciplinary collaborations among disciplinary scholars, and to foster long-term research projects that would allow this.
Another set of crosscutting solutions prioritised by the experts is improvements to the research system and environment (E). These solutions feature prominently in relation to the challenges concerning differences in research approaches, problems with the researchers' profiles, and transdisciplinary efforts in bringing research to action. The experts particularly pointed to the need for fostering more collaborative research groups and networks, and collaborative teams to enable data integration (Expert-ID25). Several experts also noted the need for creating more interdisciplinary spaces for academic exchange that for example can "enhance discussions and understanding of different languages" (Expert-ID13) and to use digital environments to build interdisciplinary collaborations in the absence of institutional structures.
The experts also focused on the need for improving the research process (D) as the solution to several scale challenges, with the main suggestions centering on the need for developing better cross-scale analytical frameworks, and methodological innovation related to upscaling and downscaling techniques. This included building better datasets, as well as linking existing datasets from the social and natural sciences since "most repositories are not interdisciplinary" (Expert-ID13). These efforts would require researchers to "set aside time and resources within a project to develop such a framework before starting the empirical data collection" (Expert-ID9). One expert also pointed to the need for adopting a relational understanding of scales to recognise "scale as an important part of research contexts that needs to be described" (Expert-ID22).
In relation to the challenges of communication and the researchers' profiles, the experts not surprisingly pointed to improvements through commitments of time and resources to discuss shared language, definitions, and concepts among collaborating researchers. Good communication skills were also emphasised by the experts highlighting the need for more open-mindedness, humility, and reflexivity among researchers regarding their own positions (C). This would allow researchers to adopt more inclusive practices and appreciate diverse research approaches, which will benefit not only the individual researcher's understanding but also the larger projects and programmes they are involved in. One expert summed up this position: "Hubris goes hand in hand with disciplinary divisions. A true researcher of wicked problems must recognise that she/he has only part of the solutions; and that her/his understanding of the problem is only partial" (Expert-ID19).
Finally, most of the proposed solutions tie into calls for more and better research funding of inter-and transdisciplinary research projects. Increasing funding was prioritised in relation to the challenges of structural problems in the academic system, problem definition and transdisciplinary projects. This was emphasised by one expert noting the need for more programmes "that contain multiple spatial and temporal scales within one common framework/research endeavour [this] takes time and money because they need to be large and long-lasting" (Expert-ID20), while another added that more funding per se is not necessarily the solution but more targeted funding would be required to address the challenges of bridging scales.

Leveraging change for bridging scales in humanenvironment research
Our analysis shows that the identified solutions are crosscutting in terms of the challenges they address. Moreover, using the leverage point framework to analyse how the different solutions can be organised according to their potential to leverage change in the academic system (Abson et al. 2017;Dorninger et al. 2020;Meadows 1999), we argue that they need to be addressed across all four types of system characteristics from the deeper level of system intent and design to the shallower levels of feedbacks and system parameters (Fig. 3).
At the deepest and most fundamental step, the solutions in the intent category correspond to changes required for affecting researchers' openness and capacity to think across scales and disciplines. Actions to enhance researchers' capability for self-reflection, humility and open-mindedness are at the basis. Through education and training programmes as well as adopting a relational definition of scale, a deep mindset change can support the conditions needed to implement the rest of the strategic solutions. Once there is an openness and understanding of the importance of bridging scales, steps towards improving the design of the academic system to institutionalise cross-scalar and interdisciplinary human-environment research can be taken. Setting principles for addressing multiple societal levels, clarifying the research focus, and developing analytical frameworks are key steps toward the institutionalisation of interdisciplinary study programmes and research approaches that consider processes and impact across scales. Then, institutional structures that facilitate spaces for academic exchange can be created.
In a next step, solutions aiming at fostering communication can be implemented through the promotion of collaborative research networks and enabling long-term collaborations. These types of internal feedbacks in the academic system would ensure actions to improve communication among researchers from different disciplines using participatory formats. In addition, they could facilitate researchers' commitment to share their concepts and definitions of scale that is necessary to improve the Fig. 3 System characteristics with the strategic solutions according to their potential to bridge scales in human-environment research. The figure is modelled on the leverage points framework presented by Meadows (1999) and adapted in Abson et al. (2017, Box 2) outreach and impacts of scientific results. Finally, more mechanistic solutions supporting the implementation of suitable analytical methods and techniques to bridge gaps or improve upscaling and downscaling data can be applied side by side with actions to increase the funding for inter-and transdisciplinary research.

Discussion
Bridging scales in human-environment research continues to challenge interdisciplinary researchers on both theoretical and practical levels. The challenges highlighted by our participants mirror the long-standing discussion around scales in the literature (Dirnböck et al. 2013;Pulver et al. 2018), including epistemological differences (Marston et al. 2005;Sayre 2005;Zhang et al. 2014), differences in terminology (Gibson et al. 2000), scale mismatches (Cumming et al. 2006), and scaling challenges (Newig and Moss 2017;Niewöhner et al. 2016;Reid et al. 2006). At the same time, the results show that these challenges are highly interconnected and that addressing them requires solutions across a wide range of interventions, from the specific to the systemic, and capacities, from the individual to the collective. Thus, bridging scales both entails an acknowledgement and better understanding of diverse scale perspectives, and the development of better methods, frameworks, and datasets that enable cross-scale analysis, as well as collective work to enable spaces that foster scale bridging at multiple levels within the academic system. As such, most of the identified solutions call for structural changes to enable shared agency from individuals and organisations, as well as contextual realisation to foster inter-and transdisciplinary education and research (Caniglia et al. 2021).

The three strategic arenas of agency to bridge scales
Based on the leverage potential of the identified solutions, we see three strategic arenas for intervention within the academic system, where change is required concerning bridging scales: (a) the individual researcher, (b) the communities, institutions, and organisations researchers collaborate and work in, and (c) the wider systemic structures of academia, including the funding bodies and framing conditions. These strategic arenas for intervention are embedded within and made up of each other. At the deepest and most foundational level, an individual commitment to openness and reflexivity, as well as to a practice of care (Care et al. 2021;Sellberg et al. 2021;Soininen et al. 2022) is required to successfully implement the solutions across systemic arenas. Yet, for enabling spaces for bridging scales both deep and shallow leverage points need to be engaged in each systemic arena of change.
In the following, we discuss how the solutions to bridge scales can be related to each of the three agency arenas, as well as the pathways to implement them. We provide examples from the literature to showcase how these solutions have been and can be operationalised and implemented in interdisciplinary human-environment research.

The self-reflexive and humble individual researcher
Individual researchers are at the core of the necessary changes required for advancing interdisciplinary research on and across scales. As individual researchers, we need to actively take responsibility for our own academic development in self-regulating learning settings (Ayish and Deveci 2019) by proactively seeking training and education to gain the necessary skills to bridge scales. This entails a commitment to self-motivated learning (Boekaerts 1988(Boekaerts , 2011 and open-mindedness about the diversity of scale concepts (Ciesielski et al. 2017). Moreover, we need a commitment to reflexivity and humbleness on the limitations of our own disciplinary backgrounds, while applying sound research approaches, theoretical frameworks, and methods (Haider et al. 2018). At the same time, we need to ensure a research focus on cross-scale interactions and work collaboratively to develop better analytical frameworks and approaches that consider both processes and impacts at different scales (Cash et al. 2006;Folke et al. 2007). Here, open communication and a commitment to discussions of shared or diverging language and terminology holds a key role as a feedback mechanism that supports such collaborative work (Wada et al. 2020). Open communication, transparent data collection, and reporting of data are also essential for addressing the challenges of bridging scales by combining and improving datasets, and existing analytical methods for upscaling and downscaling results and methodologies (Bierkens et al. 2000). Finally, we should be aware of our position within the academic system (Hernaus and Černe 2021), and work to positively influence our institutions and organisations, as well as pushing for an agenda to change the structural conditions of the academic system as elaborated on below (Care et al. 2021).

Innovative scientific institutions and organisations
The second systemic change arena, where crosscutting solutions for bridging scales need to be implemented, is within scientific institutions and organisations. For the deeper leverage points, scientific institutions and organisations need to commit to interdisciplinary education and training programmes at all academic educational levels (Annan-Diab and Molinari 2017; Boone et al. 2020). Such efforts have sprouted over the past decades, where both traditional universities and newer interdisciplinary research institutes have started recognising the need for interdisciplinary graduate and post-graduate education that teaches students to think about contemporary human-environment relations from multiple perspectives and across multiple scales. 2 In this way, scientific institutions and organisations are crucial for fostering open-minded and reflexive mindsets required for interdisciplinary human-environment researchers.
Another set of solutions relates to the design of the interdisciplinary research process and practice. Here, scientific institutions and organisations, as the providers of the necessary infrastructure for researchers to unfold cross-scale research, need to pay more attention to creating interdisciplinary spaces for exchange such as the Warwick Interdisciplinary Research Leadership Programme. This could also be in the form of resources for setting up collaborative research groups or working towards lowering barriers of collaboration between departments and researchers within institutions.
A key parameter and shallower leverage point to enable bridging scales in academic institutions and organisations is to progressively change incentive structures for highpace, quick-fix research, and instead promote long-term research projects that engage in building trustful relations among multiple stakeholders and facilitate systemic changes (Mountz et al. 2015;Paasche and Österblom 2019). Leaders of scientific institutions and organisations with agency to shape funding bodies and set the legal frameworks in academic structures should use that power to collectively aim at long-term and collaborative research that is necessary to bridge scales (Care et al. 2021;Gordon et al. 2019).

Enabling structures in the academic system
Addressing the need for fostering long-term research projects that enable cross-scale inter-and transdisciplinary research points to the need for change at a more structural level. Funding agencies and academic policy-makers have the agency to address leverage points related to both funding structures, incentive structures and evaluation criteria. One example identified by our experts is the need for more funding for methodological proposals that seek to improve existing scaling techniques and link existing datasets. Lowering barriers for sharing data within and between research organisations also presents a 'low-hanging fruit' that could foster better research across scales through using and combining existing datasets. Good first steps in this direction include collaborations within and between universities with emphasis on open access datasets, data management and stewardship adhering to the FAIR principles (Wilkinson et al. 2016), information and communication technology infrastructure, and the growth of eResearch capabilities, involving the collaboration of multi-disciplinary teams of researchers, with data scientists and computer scientists, data stewards, and digital librarians (Anandarajan and Anandarajan 2010; Steel et al. 2019;Ware and Mabe 2015). These developments to the research landscape are enabling more flexible structures for sharing and exchanging data and results.
At a more fundamental level, funding agencies, academic review boards, and promotion committees should work toward revising criteria for the evaluation and selection of research projects and researchers. Currently, emerging debates across academia demonstrate the increased attention to how the dominating 'publish-or-perish' mentality inhibits cross-scale, long-term, interdisciplinary, and experimental engagement (Davies et al. 2021).
Within the current academic landscape, however, there are some encouraging movements toward creating spaces for a flourishing practice of long-term inter-and transdisciplinary research that bridges scales (Bennett et al. 2016;Geels 2002). Examples of such developments are the Programme of Ecosystem Change and Society community, and the different platforms and umbrella organisations at the research-policy-practice interface (e.g. IPCC, IPBES, and Future Earth) that work to form new coalitions and alliances for long-term and cross-scale interdisciplinary research (Sellberg et al. 2021). In addition, more and more crossinstitutional networks for interdisciplinary collaboration are emerging. The Postdoc Academy for Transformational Leadership initiated by the Robert Bosch Stiftung 3 and hosted by four sustainability research institutes in Europe provides one example of a funding body promoting innovative space for education and exchange beyond the formally required education stages.
While simply increasing funding for long-term research and interdisciplinary education and exchange alone will not ensure a deep-seeded change in perspectives and competences to bridge scales, such initiatives exemplify the type of system parameters that should be addressed to take a first step towards enabling deeper leverage points and solutions.

Methodological reflection and limitations
The insights presented here were created by a team of researchers with different scholarly, as well as interdisciplinary training and thus a high awareness of challenges inherent to interdisciplinary research, particularly in the domain of human-environment systems. Nevertheless, designing, conducting, and writing this work in an interdisciplinary team embodied-in several regards-the challenges we identified through the study. This became especially clear in the collective coding process of the challenges and solutions into categorisations, where we encountered diverging interpretations of the statements by survey respondents. We frequently created sub-groups in varying constellations to reflect on multiple sets of results, which we compared and discussed to create robust and consistent coding schemes. Open-mindedness and jargon-free communication successfully aided in overcoming some of the challenges related to language and definitions, as did consistent protocolling and attention to good organisation and moderation of joint discussion.
We attempted by all means to avoid biases in the respondents' gender, disciplinary background, or origin. Yet, the survey participants and experts consulted in this work stem from or conduct research in 20 countries across the globe. Nevertheless, our analyses of the participant demographics revealed a strong bias towards central European countries, most notably among the survey respondents (which dominantly work in Germany). As such, this study yields a highly international representation of the research community but may not cover all issues inherent to non-European academic systems with the same depth. We thus strongly encourage further research assessing differences between academic systems across all continents, and the determinants thereof.

Conclusion
Cross-scale interactions, scale mismatches, and diverging understandings and definitions of scale continue to challenge interdisciplinary human-environment researchers in their daily work. The main aim of this study was to concretise those challenges and to scan the most suitable solutions for advancing knowledge and research practice on scales in human-environment research. The objective was thus not to "solve" the problem of scale and scaling, nor to reach a consensus on definitions. Rather, our aim was to illustrate the diversity of understandings and methods of scaling that underpins and is needed for interdisciplinary human-environment research, as well as reflect on the conditions that would enable deeper understandings and practices of scaling at different levels from individual researchers to academic structures.
The main challenges ranged from issues related to differences in research approaches targeting different scales and conflicting perspectives on problem definition, over lack of structural interdisciplinarity and constraints on resources for long-term research collaboration, to issues with communication. The identified solutions pointed to the potential of improving the research process through open communication, data sharing, or new techniques for upscaling or downscaling results. Discussing the implementation of solutions in relation to their leverage potential, we argue that bridging scales in human-environment research requires work at three strategic arenas of agency in the academic system including individual researchers, research institutions and organisations, and in the wider structures of academia. At the individual level, bridging scales requires humbleness and open dialogue between researchers from different disciplines on scale concepts and scaling methods, as well as improved collaboration for operationalising them. In parallel, research institutions and organisations should develop their capacity to enable and nurture interdisciplinary spaces to flourish, while individual researchers and institutions alike should push for a development of the wider academic system that enhances the valuation and support of interdisciplinary and cross-scale initiatives.
Author contributions All authors contributed to the conceptualisation of the study, survey preparations and analysis of the results, as well as the interpretation and visualisation of the results. CF coordinated the development of the study and led the writing of the manuscript from conception to submission. MH-M conceived the solution scanning and application of the method. All authors contributed to the writing of the manuscript and read and approved the final manuscript.
Funding Open Access funding enabled and organized by Projekt DEAL.
Data availability Data is available by request to the main author.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.