Drone imagery to create a common understanding of landscapes Landscape and Urban Planning

Negotiated solutions among contrasting land use interests in the nexus of water, energy, food and ecosystems require cooperation between actors with different viewpoints and backgrounds. We suggest aerial imagery and videos, captured by drones, to be “ boundary objects ” , easily interpretable landscape representations that might create a common understanding across stakeholders through their universal interpretability. We collected drone imagery and videos from different angles of a wide range of landscapes in Zambia, showing agricultural areas, forests, wetlands and water infrastructure. Then, we took the imagery back to the field to probe the perceptions of multiple stakeholders, including staff from both governmental and non-governmental organizations, hydropower operators, small- and large-scale farmers. In focus group discussions, we assessed the interpretability of oblique images, taken at an angle by a video drone, compared to nadir (vertical) imagery from Google Earth and from a high-end mapping drone. We show that oblique images produced better identification results across all groups of stakeholders, but especially from small-scale farmers, suggesting this type of imagery is helpful to empower people who lack previous experience in interpreting nadir images. Overall, the appreciation of the aesthetic value and the perceived professional benefits of drone imagery are high, but technical and legal barriers impede a wider adoption of the technology. While we highlight ethical concerns and technical limitations, we suggest that conservationists and environmental planners could benefit from a critical use of affordable video drones so as to produce intuitive landscape representations useful for more effective multi-stakeholder collaborations.


Introduction
The use and distribution of land and water resources is at the heart of the global environmental challenges of climate change, biodiversity loss, food security and land degradation. These global challenges fuel intensifying conflicts over land use and create trade-offs on environmental, economic, and social outcomes for various groups of people (Díaz et al., 2019;Mace et al., 2018;Newbold et al., 2015;Tittensor et al., 2014). Nature conservation and landscape planning aim to define priorities for a sustainable use of land resources, but the resulting management issues are highly complex, often referred to as wicked problems, and require the involvement of a diverse range of stakeholders (Balint, Stewart, & Desai, 2011;Defries & Nagendra, 2017). Effective tools to bridge the gap between science, policy and individual decision-making on the ground include "boundary objects" that allow bringing people with different backgrounds and capabilities together to communicate and solve conflicts (Clark et al., 2016). Boundary objects are defined as collaborative products as a base for communication that are "both adaptable to different viewpoints and at the same time robust enough to maintain identity across them" (Star, 2010;Star & Griesemer, 1989). Maps, pictures and other types of visualizations are particularly powerful boundary objects, as they are meaningful to different stakeholders and elicit their distinct contributions in participatory landscape planning processes (Ewenstein & Whyte, 2009;Schroth, Pond, & Sheppard, 2015).
While maps and aerial images have been used as boundary objects in environmental decision-making and participatory planning for a long time (Chambers, 2006), dronesor unoccupied aerial vehicles (UAV, Joyce, Anderson, & Bartolo, 2021) have the potential to improve the accessibility of aerial views to decision-makers and the wider public (Kullmann, 2018). Participatory mapping approaches have been influencing land use planning increasingly (Eilola, Käyhkö, Ferdinands, & Fagerholm, 2019;Jaligot, Hasler, & Chenal, 2019). Yet, the main limitation for interpreting maps and satellite images in participatory processes remains that not everybody gets the same information out of those, and their use and production is not equally accessible to all people (Brown & Fagerholm, 2015;Carolan, 2009). While drones share some of these limitations (Davies, Lai, & Chua, 2018), they have diversified the toolbox of visual geospatial materials through videos and still imagery, at multiple angles and at very high resolutions, which allow "a level of clarity comparable to the world that we perceive from the ground" (Kullmann, 2017). The film industry has already embraced the power of this technology to provide bird's-eye-views to capture the viewer's attention and provide unusual viewpoints (Zacharek, 2018). The advantage of aerial filming is that it might "extend a perceived territory towards a non-anthropocentric vision" and thus "de-territorialize the human ways of looking" (Mikkola, 2020, p. 202). Filming a landscape from the air used to be very expensive, because it involved a plane or helicopter, but it has now become affordable to many more users through the availability of cheap, yet sophisticated, civilian drone technology for both recreational and commercial purposes (Munck Petersen, 2020).
The use of drones in environmental sciences and conservation has become widespread, for example for surveillance, mapping and species detection (Wich & Koh, 2018), agriculture and land use monitoring (Librán-Embid, Klaus, Tscharntke, & Grass, 2020) and hydrological assessments (Samboko et al., 2020). Despite the wide availability of drones and the new opportunities that provide compelling new perspectives at an affordable cost, the use of the technology is dominated by a small, relatively homogenous group of people (Rogers, Singh, Mathews, & Cummings, 2022), which has led to multiple initiatives teaching drone technology to wider parts of society, including indigenous peoples Joyce, Meiklejohn, & Mead, 2020;Paneque-Gálvez, Vargas-Ramírez, Napoletano, & Cummings, 2017). Yet, strong reservations against using drones still prevail amongst policymakers and society, likely due to the military origin of the technology and increasing misuses that lead to a perception of invasiveness (Harriss, 2020). Issues related to the use of drones, such as safety, privacy, psychological wellbeing and data security, need to be taken seriously (Sandbrook, 2015), while the legal and administrative frameworks for a safe and regulated drone use of are still developing. In large parts of the world, drone regulations remain restrictive and change frequently, making it difficult for users to stay updated on the latest legal situation (Stöcker, Bennett, Nex, Gerke, & Zevenbergen, 2017). In a recent survey for Africa, for example, 14 countries did not have any UAV regulations, two had a complete ban, and for many others the regulations were highly untransparent and changing (Haula & Agbozo, 2020). Given this background, it is necessary to critically reflect on the potential, actual uses, and barriers of mainstreaming drone imagery as a tool for landscape planning. It is particularly important to take into account how various stakeholders perceive such imagery in the context of rapidly changing landscapes in the tropics.
In this paper, we explore the potential of drone-based imagery for collaborative land management. Based on focus group discussions with a wide range of stakeholders in Zambia, we assessed how different types of drone imagery facilitated the intuitive interpretation of landscape elements across various groups of stakeholders who manage land as farmers or administrators. We further assembled qualitative information on the potential, actual uses and perceived benefits of drone imagery and on potential and experienced barriers to use drones. We reflect on our experience and propose a way forward by illustrating which type of visual materials should be collected to allow developing joint landscape visions, and to determine the enabling conditions from a technical, administrative, and ethical point of view to influence the sustainability of land and water use.

Drone data collection
We conducted this study in the Zambian part of the Zambezi River Basin in Southern Africa, representative for a region where the populations and economy are growing rapidly and irrigated agriculture and hydropower production are expanding rapidly. As people's demand for water and land is competing with the environmental functions and services (World Bank, 2010), finding compromising solutions between the different landscape users is crucial, with transboundary organisations such as the Zambezi Watercourse Commission (ZAMCOM) providing learning spaces to negotiate compromising solutions between the actors across various sectors and national borders (Lumosi, Pahl-Wostl, & Scholz, 2019).
We collected UAV imagery from seven different landscape settings during two field trips in March and September 2018 (Fig. 1). Together with researchers and students from the University of Zambia, landowners and administrators, we purposively selected appropriate sites so as to highlight contrasting agricultural and natural landscape features of importance for the water, energy, food, ecosystem nexus in Zambia (Winton et al., 2021), including forests, rivers, wetlands, small-, and large-scale cropping systems. All flights were carried out by a licensed drone pilot with permission from the Zambian Civil Aviation Authority and supervised by a member of the Zambian Air Force. At all sites, we collected video and still imagery. For the videos taken from different angles, we used a quadcopter-type Mavic Pro (DJI, Shenzhen, China). For the still images at nadir angle, we used a fixed-wing Ebee, equipped with a Sensor Optimised for Drone Applications (SODA) photogrammetry camera with a ground resolution of 2-3 cm per pixel (Sensefly, Geneva, Switzerland).
Drones enable new shot-types and moves in cinematography, which capture the landscape perspectives from multiple viewpoints (Mademlis et al., 2019). In general, the type of imagery collected by aerial sensors differs depending on the angle of the optical axis. The three main directions ( Fig. 2) are called "nadir" (90 • vertical angle), "high oblique" (<90 • and >30 • ) and "low oblique" (<30 • and >0 • ) (Sheikh, Khan, Shah, & Cannata, 2003). The nadir images are spatially the least distorted which makes these the most applicable for quantitative spatial analysis, however, images with oblique perspectives have also been in use since the beginnings of photogrammetry (Petrie & Nolan, 2018;Vacca, Dessì, & Sacco, 2017). The main difference between the viewing angles is the scale accuracy, i.e., the measurable distance between objects on the ground, which increases from low oblique over high oblique to nadir.

Focus group discussions with stakeholders
In November 2021, we met with nine groups (each consisting of 2-10 people) and three individual stakeholders in Zambia so as to conduct focus group discussions based on semi-structured interview questions and an interactive image classification exercise (Fig. 3). Eight groups were selected because they manage all, or parts of, the land covered by the drone imagery collected three years earlier. Four more groups had an administrative relation with the areas as part of governmental, nongovernmental or semi-private organizations. Overall, we included water administrators, hydropower operators, an environmental NGO (four groups of the "Admin/NGO" category), local residents of rural areas practising small-scale and subsistence agriculture (four groups of "small-scale farmers", owning <5 ha of land) and owners or managers of large-scale farms (four groups of "large-scale farmers" owning >50 ha of land). One large-scale farmer was at the same time managing a large, protected forest area. In total 45 people participated in the focus group discussions (32 men, 13 women) with a wide range of educational and professional backgrounds. The discussions with small-scale farmers were held in Tonga and Bemba with key points translated to English. All  other discussions were in English. We guaranteed anonymity to all participants as part of the ethics protocol (EK 2021-N-180, ETH Zurich Ethics Commission) and did not request any personal information. Pictures were only taken with written consent.

Comparative image classification
One part of the focus group discussions consisted of a landscape element classification exercise. In three iterative steps, we showed printed aerial images of an area near Kafue town based on 1) Google Earth imagery (https://www.maps.google.com), 2) Nadir drone imagery (captured by Ebee) and 3) high oblique drone imagery (captured by Mavic Pro). The original images are available as supplementary information (SI). For each image, we requested the stakeholder groups to classify landscape elements and to label those according to their interpretation. None of the participants knew the area, except one group of small-scale farmers who were living exactly there and who are therefore referred to as "local residents" in the result section. Many of the smallscale farmers were illiterate, but still participated in the exercise while instructing a younger family member to write down the labels. In the end, we counted the number of labelled items. We did not compare the accuracy of the labels across the groups, as these were often not at a comparable abstraction level. For example, some groups would use the general label "vegetation", whereas others would specify "papyrus" as the specific vegetation type while others would use a local term for which there might not be a precise English translation. The purpose of this exercise was to assess which type of imagery was the easiest to interpret rather than to evaluate how proficient participants were at classifying various types of land cover. We therefore used the number of uniquely labelled items as an indicator for image interpretability.
We used analysis of variance (ANOVA) to compare the classification results in-terms of interpretability between the three types of imagery across all stakeholder groups. A normal distribution was confirmed by both the Levene and Shapiro-Wilk tests. We used the Tukey multiple pairwise-comparisons to see which image type exactly differed significantly from the other. The number of samples was n = 12 for each image type, distributed in four stakeholder categories: a) four groups of staff from government agencies or NGOs (Admin/NGO), b) four groups of large-scale farmers, c) four groups of small-scale farmers, of which d) one group included the local residents. Given the small sample number and the purposive sampling strategy, the results are clearly not representative for a well-defined population. We rather aimed to provide a proof of concept based on a qualitative analysis of the drone imagery perception.

Limits and opportunities of drone use
During the focus group discussions with stakeholders, we clarified the differences between the various types of drones and the imagery they can produce, and we then asked the groups about their perceived opportunities and barriers to use drone imagery. Each group was provided with drone-based images printed at poster size and taken at nadir as well as at oblique angles, of the area the respective group was familiar with (small extracts and examples in Fig. 4).
We then asked open-ended questions on the current and potential future use of drone imagery in comparison with other spatial data and on the barriers to making wider use of the technology. The interviews and focus group discussions were recorded with the consent of all participants. Based on these recordings, we extracted and coded information on whether or not the group is currently using, or has used drones in the past, and listed the potential and actual uses and benefits, as well as potential and experienced risks and limitations by each group. We coded the responses thematically and selected which theme had been mentioned by which group, in order to identify potential patterns in the responses depending on the stakeholder category.

Image classification
The image classification exercise showed significant differences in the number of identified landscape elements between the three image types, across all stakeholder categories (three-way Anova P = 0.001). Overall, the number of identified classes (Fig. 5 B) varied between 8.83 for Google Earth (±1.71 95 %-confidence interval), 11 for the Nadir drone image (±1.93) and 14 for the Oblique drone image (±2.31). The pairwise Tukey test indicated a significant difference in interpretability between Google Earth and Oblique drone imagery (P < 0.001) but not between Google Earth and Nadir drone (P = 0.226) nor between Nadir drone and Oblique drone (P = 0.065). The improvement in image interpretability separated by stakeholder category (Fig. 5A) was most pronounced for the small-scale farmers (median of 6 classes identified from Google Earth, 8 from the nadir drone and 12 from the oblique drone imagery) when compared to the large-scale farmers (10, 12, 12.5 respectively) and the Admin/NGO groups (9, 11.5, 12.5). That means, the groups of small-scale farmers only identified half the number of classes as compared to the large-scale farmers and the Admin/NGOs when using Google Earth and nadir drone images. Yet, this difference in interpretability of the type of imagery did not exist when using oblique drone imagery. The small-scale farmers, who were residents in the area depicted in the imagery, identified more classes than any of the other groups of stakeholders (11, 16, 23) as they could add detailed information about the uses of the different landscape elements. In general, all groups clearly preferred the drone imagery over Google Earth, as expressed by the comment "it feels like taking my glasses on and off" (public administration employee). The small-scale farmers, as well as many of the other groups, expressed their obvious preference for the oblique drone image over the nadir perspective: "You start thinking

Use and perception of drone imagery
The focus group discussions show that drones play a minor role in the professional practice of our focal Zambian stakeholder groups (Table 1). Although three groups of Admin/NGO and two large-scale farmers have already used drones occasionally, they stated that their importance is lower than other sources of spatial data, especially Google Earth and other satellite imagery. The actual experiences with drones are limited to recreational uses or the inspection of agricultural land and water infrastructure in real-time. As one large-scale farmer mentioned: "We use it occasionally on the farm to fly over the field to check whether or not the tractor is still working". In one case, drone images have been used to complement satellite-based remote sensing analyses and in one other case, an energy producer collected drone imagery to show the dam infrastructure to the wider public. Despite the currently limited use, when asked for the potential future use and benefits, there was no shortage of ideas. An often-mentioned need was the surveillance and control of illegal activities, including agricultural encroachment, water abstraction, poaching, and charcoal production. Further potential applications include visualizing boundaries of land and property, monitoring changes over time, game counts for conservation and, as mentioned by all large-scale farmers, the support of precision agriculture, e.g. to inform targeted use of fertilizers and pesticides.
One benefit almost every group mentioned, after looking at the example of drone imagery, had to do with the beauty of being able to see landscapes from above at a very high resolution. This is linked to the ability to contextualize inhabited and cultivated areas together with neighbouring areas that cannot easily be reached, such as wetlands. The employee of an NGO commented on the printed images: "You can put a frame around this and put it in your house" and a large-scale farmer said that "It is much more exciting to see the whole landscape from above, as we are used to from TV shows and movies". Some people explicitly mentioned the potential usefulness of such imagery as a communication tool between stakeholders, e.g. for conflict resolution in land disputes. Two of the Admin/NGO groups also mentioned the value of such images to communicate with the wider public: "If you can show people, look, this is the area covered by an invasive species, and this is how the wetland used to be. Based on an image like this, it already rings a bell".
Although potential uses of drone imagery were mostly based on anecdotes and the imagination of stakeholders, the barriers limiting use of drones are more concrete. Stakeholders describe official regulations to be too strict or simply unclear. Laws are missing that differentiate drone types and purposes, as indicated by the following quotes: "There is a need to train the aviation authority themselves to be able to make a difference between different types of drones" (NGO staff), and "We do not use drones at all because the regulations are too strict. We know that many other farms use them on their private property, but we do not want to do anything that is against the law" (Large-scale farm manager). Other constraints mentioned include the scarcity of skilled drone pilots, the time required for image analysis, the cost of hard-and software, and the storage of imagery. Some groups also mentioned practical limitations such as the sensitivity of the drones to weather conditions, the difficulties in finding sites for take-off and landing, as well as the limited reach of drones and, consequently, the area that could be covered at once. Additionally, all of the small-scale farmer groups, none of whom had any direct experience with drones, mentioned some risks regarding the actual capability of drones.. They had particular reservations relating to being observed and hence potentially losing their land, which is a pertinent fear in a situation of insecure land tenure. These worries were, for example, expressed as "We are concerned, because you never know if someone will come and try to take land away from us", or "It feels like someone is trespassing onto our property".

Discussion
We encountered strong asymmetries in the perusal and interpretability of various types of drone images between categories of stakeholders. While all stakeholders identified more landscape elements when using oblique drone imagery, the ones who benefitted the most were the small-scale farmers, who have the least formal education, wealth, and power in the Zambian context of high income inequality (IGC Zambia, 2017). Most employees of the hydropower companies, public administration, environmental NGOs and the large-scale farmers enjoyed higher education and already had some experience with maps and spatial data, which seemed to help with the interpretation of the satellite and drone-based nadir imagery. In contrast, most small-scale farmers did not know how to read or write, reflecting the lower level of formal education of people living in rural areas in Zambia (Chapoto, Subakanya, Beaver, & Kuteya, 2019). Yet, the oblique imagery allowed all stakeholders independent of their backgrounds to have a similar, intuitive understanding of the elements on the ground and their spatial configuration. Oblique aerial images have been shown before to be a powerful tool to enable effective landscape communication (Svenningsen, Brandt, Christensen, Dahl, & Dupont, 2015). The increased availability of drone visuals allows more people from various backgrounds to benefit from such new viewpoints (Hagan, 2021). Hence, drone images as "boundary objects" can help reduce the disparities in information and understanding. In a world where information means power, access to drone imagery can empower people from socially more vulnerable groups. Our results provide evidence for previously suggested uses of drones for empowering local governance and enabling knowledge sharing for co-development of land stewardship (Macdonald et al., 2021). The oblique images, as stills and videos, seem to be very suitable for such uses as shown by our classification exercise. Each viewer interprets an image based on their background knowledge, experiences, epistemological frames and strategic interests. The bird'seye-views captured by drones have the advantage to offer spatial representations of the landscape, which could be almost universally understood.

Building joint landscape visions
Drones have an important advantage as they allow the collection of Table 1 Perceived benefits and limitations of using drones in the practice of different categories of stakeholders. Each column represents one focus group discussion, each row indicates a mentioned subject of potential (O) and experienced (X) uses and benefits, risks and limitations. Two groups of large-scale farmers were interviewed jointly, which explains why there are only three focus groups as compared to four in the image interpretation exercise (Fig. 5).

Sensitive to wind and weather conditions
-

Risks
General skepticism because purpose is not clear both oblique and nadir imagery, which could fulfil many of the uses desired by stakeholders, such as monitoring land cover changes, assisting in the demarcation or surveillance of boundaries, and improvements to agricultural applications. The stakeholders frequently mentioned the need for real-time surveillance and interpretation of what is happening on their land and further highlighted the need to show "their" landscapes to policymakers and the wider public. Here, drones complement the toolbox of maps, aerial photographs, satellite images, and Geographical Information Systems (GIS) that have since long been part of tools applied in participatory land management (Chambers, 2006;Harvey & Chrisman, 1998;Slocum, Wichhart, & Rocheleau, 1995).
A key component of our research was to bring drone imagery back to the area where we had collected it and to discuss and assess the perceptions of people living and working in the area who may not regularly work with such data. In addition to their widespread support and interest, the stakeholders also expressed the expectation that, despite the current technological and legal barriers, drones could improve their daily practices. We found that the aesthetic value of drone imagery helps with the appreciation of the technology and may play an important role in the wider adoption and acceptance. The oblique images, taken by a relatively cheap Mavic Pro (video) drone, allow the best representation of a landscape in terms of interpretability by a wide range of people. So far, the published examples of drone-based participatory planning and mapping approaches are largely based on nadir-angle imagery. The examples include the development of village maps for participatory regional planning (Nurdin et al., 2019) and the determination of cadastral boundaries as a low-cost solution with community involvement in Indonesia (Ramadhani, Bennett, & Nex, 2018). Further, interviews, assisted with drone imagery, with farmers on the Galapagos Islands did not only reveal visible spaces but also management strategies and agrarian knowledge (Colloredo-Mansfeld, Laso, & Arce-Nazario, 2020). Based on our study, it would be hard to justify the deployment of expensive, professional mapping drones, but video drones may provide an affordable and useful tool that can benefit stakeholders by providing detailed representations of the landscape.

Issues and limitations
The focus group discussions and our own first-hand experiences with drone-based data collection in Zambia have revealed the existence of strong barriers preventing the more widespread use of the technology. This finding is in line with the previously reported difficulties while collecting drone imagery due to the technical, administrative, and ethical issues and limitations (J. P. . The technical limitations depend on the used drone type, with the capacity of the battery being the most important determinant of the flight range and duration. Flight planning in unknown terrain can be challenging, as appropriate take-off and landing sites close enough to the area of interest might be hard to find. Furthermore, the weather conditions can pose other constraints. Rain, wind and strong sun can make a safe and effective operation of a drone impossible. Moreover, it can be difficult to comply with all of the regulations as they are often vague and subject to changes. In the Zambian case, this applies both to the importation of material and the acquisition of flight permission. For example, every single flight had to be authorized and supervised by a liaison officer from the Zambian Civil Aviation Authority. The focus group discussions further showed that an important perceived limitation to the use of drone imagery is the need for technical training. Some of our drone images were collected with staff and students from the University of Zambia in Lusaka during a one-week training workshop. The workshop included the planning of suitable areas for data collection, obtaining permits, flight planning, identification of suitable take-off and landing sites in the field, manual and autopilot flight control, data management and visual interpretation of the imagery. Other similar workshops also included the construction of lowcost DIY (Do It Yourself) drones (Paneque-Gálvez et al., 2017). Yet, in order to see and peruse drone imagery, one does not need to have, or to know how to fly, a drone. People around the world are sharing their pictures and videos on platforms like Dronestagram (Hochmair & Zielstra, 2015) and through the open access platform OpenAerialMap (Johnson, Ricker, & Harrison, 2017).
The discussions with the small-scale farmers, who do not currently have access to drone technology, highlight important ethical considerations. Flying a drone over their land can be intimidating if the purpose is not clearly explained, and the imagery produced can be extremely powerful to those who control the process of flying, filming, processing, and interpreting. In a situation of insecure land tenure, and in an environment where land ownership is frequently challenged and changed (including so-called "land grabs") (Manda, Tallontire, & Dougill, 2019), which is common in Zambian smallholder systems (Akayombokwa & Van Koppen, 2015), drones could seem like a real threat to people's livelihoods.
When conservationists and planners use drones to document landscapes, they should avoid having drone technology reproduce stereotypical images of human-nature interactions or become part of the militarization of conservation (Duffy, Massé, Smidt, Marijnen, Büscher, Verweijen, Ramutsindela, Simlai, Joanny, & Lunstrum, 2019). Such a use could lead to an atmosphere of fear and alienation of the local populations (Humle, Duffy, Roberts, & John, 2014). Especially in urban areas, the privacy and land ownership need to be respected, as drones could potentially depict sensitive areas such as backyards that are not normally open to the views of strangers (Gevaert, Sliuzas, Persello, & Vosselman, 2018). While it may seem useful to share the imagery publicly, ethical concerns with online publishing without consent and critical review remain, as it may expose people, their livelihoods, and privacy to the public eye (Stankov et al., 2019). Access to the technology itself is unequally distributed, globally and within countries, giving a strong responsibility to those who have access to it.

Conclusions
The wide availability of drone imagery potentially heralds a revolution in participatory landscape analysis, as it enables more people than ever before to see their surroundings from a bird's-eye view. The popular media have embraced the visual power of drone video and the private and commercial uses are booming. Despite the technical issues and ethical concerns, conservation and environmental planning should not shy away from making good and critical use of such imagery to inform and stimulate participatory processes. The results of the focus group discussions show that using oblique drone imagery instead of nadir imagery aids all stakeholders in better understanding spatial information. Our research provides a proof of concept for further studies on the interpretability of different types of drone imagery across people with different backgrounds and in different regions of the world. Future studies could establish a more quantitative approach, with higher numbers of participants and a randomized sampling strategy. This would provide further evidence of how drone imagery can provide better information utility for a wider array of stakeholders as compared to other types of spatial information. Our results show a clear direction, namely that if transparent regulations and ethical safeguards are in place, the use of entry-level video drones might help to create a common understanding amongst landscape users across backgrounds, sectors, and interests and thus help solve complex land management issues.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
I have shared a link to my data in the manuscript. All data and interpretation results from the focus group discussions are made available in anonymized form through the ETH Research Collection, under DOI https://doi.org/10.3929/ethz-b-000566454. Nadir still imagery is freely available through OpenAerialMap (https://zenodo.org/record/ 4338899#.YxxdDTWxWUk).