Citizen science and geomorphology: the citizenMorph pilot system for observing and reporting data on landforms

Contributory citizen science projects face challenges regarding data quantity and quality. To counteract this, the projects must be centred around citizen needs and preferences, while considering aspects such as the data contribution process, including instructions, project promotion, information provision, feedback and recognition, and the design of the respective elements. Based on an understanding of the relevance of these issues affecting data contribution systems, we must determine which elements we can use to meet citizens’ needs and preferences and how to better tailor the system design to citizens’ requirements. The citizenMorph project, which aimed to create a pilot system for citizens to collect and report data on landforms, focused on the development of a citizen-centric system with elements that foster and encourage citizen engagement. We used a specifically conceived development workflow that combined participatory design with the prototyping model to involve citizen representatives in different ways and to different degrees in requirement specification, system design and implementation, and testing. This allowed citizens’ requirements to be specified and comprehensively considered in the citizenMorph system. Based on the input of citizens who were involved in the development process, the citizenMorph pilot system includes a data contribution application and a project-related website with several project-specific elements that focus on attracting and recruiting citizens to participate and increase their initial and ongoing engagement and willingness to report landform data. This includes traditional and web-based promotion elements, a specifically designed information strategy that considers information detail, depth and presentation media, project and task-tailored data contribution instructions and support, and the possibility for users to find and view the data they contributed on a web map.


Introduction
Citizen science has been used in many domains, including those concerned with environmental issues (Schaffer andTham 2020, Peter et al 2021). Of the different forms of citizen science, i.e. contractual, contributory, collaborative, co-created, and collegial (Shirk et al 2012), contributory projects constitute the majority (Bonney et al 2009a, Kermish-Allen et al 2017, Fischer et al 2021, Land-Zandstra et al 2021. Contributory projects include passive sensing, crowdsourcing, environmental/ ecological observations, and participatory sensing (Shirk et al 2012, Haklay 2021. Advances in information and communication technologies have created new possibilities for contributory citizen science, i.e. to integrate the public, their knowledge, resources, and commitment to scientific research (Bonney et al 2009b, Fischer et al 2021, Peter et al 2021. This includes using geospatial technologies to allow citizens to generate and report spatial data (Haklay 2021, Skarlatidou and Haklay 2021). Such contributory projects are thus related to the concept of voluntary geographical information (Haworth et al 2018, Haklay 2021, which describes the use of online platforms to involve citizens in the voluntary creation, assembly, and dissemination of spatial data (Goodchild 2007, Brown andKyttä 2014), whereby data contribution is based on mapping activities or implicitly associating data with geographic locations through geotagged photographs or microblogs (Craglia et al 2012, Senaratne et al 2016).
Several benefits are associated with contributory citizen science: provision of data that are otherwise unavailable or difficult to obtain, such as people's observations, their local spatial knowledge, and how they perceive, experience, value, and use infrastructure or resources (IFAD 2009, Brown and Kyttä 2014, Herfort et al 2015. Data generated in the context of citizen science can cover a large geographic scale and a long period (i.e. to support monitoring), which is often challenging for traditional data collection approaches because of staff and budget constraints (Cohn 2008, Herfort et al 2015. Citizen participation promotes learning and acquisition of scientific and digital skills and knowledge related to the topic under consideration (Hecker et al 2019). However, contributory citizen science also faces challenges. The number of citizens willing to contribute to an initiative is often less than expected, citizens can be difficult to reach and engage with, or they close the application without contributing much or any data (Nielsen 2006, King and Brown 2007, Crowston and Fagnot 2008, Morais et al 2013. Another disputed topic is data quality, which covers, for instance, precision and accuracy, validity and reliability, completeness, consistency, timeliness, fitness for intended use, and bias (Balázs et al 2021, Downs et al 2021. Depending on the number of contributors, their level of domainrelated knowledge, digital skills, and motivation, the quality of their contributed data can vary (Bonney et al 2009b, Haklay 2013. To fully exploit the potential of contributory citizen science projects and to address the challenges of ensuring that data are collected and reported in sufficient quantity and quality, applications must be centred around the users (i.e. citizens) and their needs and preferences (Haltofová 2020, Hennig 2020, Silva 2020). Methods and tools are required to promote projects because people can only contribute to a project they know about (King andBrown 2007, Schaffer andTham 2020). Relevant project-related information (Butt 2015), including support for data contribution activities (Schaffer and Tham 2020), must be adequately conveyed. Feedback and recognition also play a key role since their absence results in citizens dropping out of a project (Veeckman et al 2019, Fischer et al 2021. Applications must be designed to make them attractive, easy and comfortable to access, intuitive, and fun to use (Newman et al 2010, Mooney et al 2016, Fischer et al 2021. These aspects, which are also considered success factors for contributory initiatives, should be considered when developing systems that enable data contribution in the context of citizen science (Asingizwe et al 2020, Haltofová 2020, Hennig 2020, Schaffer and Tham 2020). A combination of different components, elements, and materials (i.e. building a system) is thus relevant for contribution to citizen science projects. The question is, which elements can we use to actually meet citizens' needs and preferences and how can we better tailor the system design to their requirements? In the citizenMorph project (Observation and Reporting of Landscape Dynamics by Citizens; 2018-2020; https://citizenMorph.sbg.acat/), we aimed to address these issues for a geomorphology-related application. The focus was on the development of a citizen-centric system, in which the data (contributed by citizens) is used in further analysis steps.

The citizenMorph project
The citizenMorph project aimed to develop a pilot system that allows citizens to collect and contribute field data on landforms (location, landform type, geotagged photo for the overview of the landforms observed, and collection of images for multiview 3D reconstruction). The intention was to add a citizen science component to a larger research project (MORPH-Mapping, Monitoring and Modelling the Spatio-Temporal Dynamics of Land Surface Morphology; 2016-2020), which focused on studying natural hazards and geomorphological features using remote sensing and enrich and broaden its outcomes with citizens' field data (figure 1).
Although novel methods for analysing the spatialtemporal dynamics of surface morphology by integrating various optical and radar remote sensing data have recently been developed (Dabiri et al 2020), there is still a high demand for data gathered directly in the field (e.g. for reporting and describing recent geomorphological events, landform characteristics, and landscape changes). Field data are of great importance for enriching and validating aerial and satellite remote sensing-based mapping results, as well as for increasing their detail and information content. Data collected in the field can aid spatial accuracy assessments, support change detection analyses, reveal unknown landscape information, and pinpoint details that are not visible in remote sensing imagery due to, for example, limited spatial resolution or the viewing angle. For example, small glacial features, such as hummocks, might not be detectable on remote sensing imagery but can be recognised in the field. Field photographs can also be used for multiview 3D reconstruction of the surface using structure from motion (SfM) and dense image matching methods (Remondino et al 2014, Brush et al 2019. Thus, high-quality field data and remote sensing data are important for comprehensive analysis and broadening our knowledge about geomorphological landscape dynamics and the prevalence of landforms (Napieralski et al 2013). The integration of (multitemporal) data from different sources facilitates the creation of more detailed and up-to-date geomorphological maps, supports studies on landscape evolution, and increases our understanding of recent landscape changes caused by natural hazards and processes.

Workflow and development approaches
The development of the citizenMorph system relied on the combined use of participatory design and the prototyping model. Participatory design involves users (referred to as 'citizen representatives' in the following) in all aspects of an application development process, such as requirement specification, design, implementation, and product testing (Baek et al 2007, Majid et al 2010, Wilkinson and De Angeli 2014. The close cooperation with the future users of a product gives us a better understanding of their requirements and skills and enables us to produce applications that allow users to achieve outcomes in a better way (Steen et al 2007, Muller andDruin 2012).
The prototyping model is an evolutionary software development process model. It is particularly suitable when the requirements of future users of a product are not well understood at the beginning of a development process or when developers are dealing with an unknown or new topic. The iterative nature of this model is crucial. Based on an initial specification, the product requirements are continuously developed and improved until the users are satisfied. This relies on an iterative loop of design, implementation, and discussion of prototypes. Then, the final requirements are specified and form the basis for the design, implementation, and testing of the final product (  Combining participatory design and the prototyping model, the workflow for the development of the citizenMorph system comprised four steps (figure 2): (a) specification of citizens' initial requirements, (b) development and discussion of prototypes by citizens, with further specification of citizens' requirements, (c) system design and implementation, and (d) testing. citizens were involved in these steps to varying degrees and in numerous ways.

Work steps
Step 1: specification of citizens' initial requirements An online survey was conducted to gain insight into citizens' needs and backgrounds. Created with Sur-veyMonkey, it contains open and closed questions covering different items in three groups (table 1; https://de.surveymonkey.com/r/citizenMorph). Cooperation with citizen representatives in its creation ensured that the questionnaire was tailored to citizens' needs (e.g. avoidance of technical terms). The survey was conducted in German-speaking countries between August and December 2018 and promoted through e-mail, newsletters, social media, and word of mouth. The data provided by the questionnaire were statistically analysed. A quantitative content analysis was used to analyze open questions.

2.3.2.
Step 2: prototyping Citizens' contribution to prototyping took place between winter 2018 and summer 2019. After an introduction to methods and tools, citizens in small groups developed prototypes that corresponded to their needs and ideas for recording and reporting landforms. During four workshops and independent group work, the prototypes were implemented using ESRI's Survey123 for ArcGIS Online. They were discussed by citizens and scientists, and additional citizen requirements were identified.

Step 3: design and implementation
Taking the citizens' needs (step 1, step 2) into account, the data contribution application and related materials were developed (e.g. information provision, instructions, giving feedback and recognition). The data contribution application was created by the scientists using ESRI's Survey123. The related material was created in different ways, by the scientists, citizen representatives, or by scientists and citizen representatives together (summer 2018 and 2019). This relied on the use of the content management system (CMS) WordPress, ESRI products, and different multimedia tools. Ethical concerns (e.g. handling of personal data, intellectual property rights) were discussed and approved by the Legal Department and the Data Protection Coordinator of the University of Salzburg (Austria).

2.3.4.
Step 4: testing Citizens tested the pilot system for acceptability, usability, and accessibility during excursions to (a) the German Berchtesgaden National Park and Austrian Weißbach Nature Park (11 July 2019) and (b) Hofn, Iceland (5 September 2019). After contributing landform data using the citizenMorph system during their field trips, participants reported on the citizen-Morph system and its use in subsequent group discussions, see Downey (2007) for more information about group usability testing.

Citizen involvement
Numerous citizen representatives from diverse backgrounds, including special needs users such as youth and the elderly (Kirakowski et al 2003), took part in the development process. This allows the creation of more useful and accessible solutions that benefit all users, which is particularly important for applications aimed at the general public a heterogeneous group of people with different backgrounds, knowledge, and requirements Tobón 2003, Steinmann et al 2004). Special needs users play a relevant role in application development since they are very demanding, critical, and sensitive to usability and accessibility issues (Brown and Hollier 2015, Gottwald et al 2016,

Questionnaire results
Hundred and thirty eight people participated in the online questionnaire intended to specify citizen's initial requirements. Selected results are shown in the appendix. Sociodemographic aspects of the questionnaire results are shown in figure 10. Most respondents had not heard the term 'citizen science' before (table  A1), and only a few had previously participated in citizen science projects. The quantitative content analysis revealed that what the citizens liked best about the citizen science projects was that they could contribute, learn something new, and collaborate with others. None of the citizen science project examples provided related to geomorphology. Only a third of the participants declared interest in the topic of geomorphology for citizen science projects (table A1).
In terms of their preferences for a citizen science application (table A2), most respondents indicated that their preferred means for receiving information is text with pictures, and videos. Enough background on the project and clearly stated aims and objectives were considered important, and the option of face-to-face contact with others was also highlighted. Previous experience contributing to online maps was limited (20% contributed spatial data or performed map editing; table A3), but the majority were familiar with online maps for route planning, orientation, and address searches. Almost half of the respondents said they enjoy working with online maps and spatial data.

Prototyping
Three prototypes were jointly developed with citizen representatives (senior citizens). The discussion rounds focused on the main topics highlighted in section 1: (a) project promotion, (b) the type of information provided, including support and instructions, (c) the data contribution process and the design of the application, and (d) feedback and recognition (figure 3).

The citizenMorph system
Based on the citizen's requirements (i.e. questionnaire findings, prototype discussions), the citizenMorph system consists of two interlinked web-based components (figure 4): a data contribution application and a project-related CMS website. Web-based methods and tools as well as traditional analogue methods and materials were used to inform and recruit citizens to participate in the project.

Data contribution application
The data contribution application was developed as a browser-based Survey123 smart form, which is an online questionnaire with information and instructions. The application allows users to contribute (spatial) data on landforms on-site or at home using different devices, such as mobile devices or desktop computers. The application can be used without registration or login. After a short welcome, the citizenMorph project is introduced, whereby participants can choose between written information, a video, or different audio files. All information is kept short and to the point, irrespective of the chosen media. Further information is available on the project website and can be accessed through links from the application.
The application provides different questions for participants to report on landforms ( figure 5). This includes a map-based question to locate and map the landform and two image questions to upload photos of the landform (overview photo and photo series). The use of branch-logic (i.e. subsequent questions are only presented according to previous responses) supports citizens in categorizing landforms. While completing the survey, citizens can directly enter the type of landform (if known) or be guided. If guided, different landform categories are presented. Once the category is chosen, the associated landforms are displayed for selection. Only the 'consent of use' question is mandatory. No personal data is collected. Participants can optionally provide an e-mail address, used to send additional information about the project. To guarantee that no personal data is collected, participants are requested to add only e-mail addresses that do not include their first name and last name, and company name (i.e. first_name.last_name@company_name.com; EC 2022). The e-mail addresses are checked and deleted if they do not correspond to these requirements.
Although the application was created to be as self-explanatory as possible, instructions on landform categorization, mapping, and taking photos are available. Clear and easy to follow instructions were created with the support of the citizen representatives. Particular attention was paid to avoiding technical terms. Instructions are optionally available with varying degrees of detail and depth, both in written text and audio files (figure 6). The audio files (approximately one minute) were implemented in the form of a question-and-answer dialogue.
For landform identification and categorization (i.e. branch-logic), citizens are supported through specifically created landform factsheets that include a short and easy-to-understand description of the Thus far, 28 factsheets have been created (figure 8), grouped into five landform categories: mass movement related, glacieal, volcanic, and coastal features, and other landforms. The landform classification corresponds to approaches commonly used in geomorphology (Huggett 2011). Additional landform categories and landforms (including factsheets) can easily be added.

Project-related CMS website
The CMS website contains comprehensive information on different project-relevant aspects and complements the content of the data contribution application. It includes project information, i.e. aims, relevance of the topic (from the citizens' point of view), domains behind the project (geomorphology, geoinformatics, remote sensing, and citizen science), and the project team, with contact information. The relevance of the field of geomorphology to the daily lives of citizens and the topicality of landscape dynamics regarding natural hazards and their connection to climate change are emphasised and explained. This also includes details about how the data is used. In addition to the data contribution instructions provided in the application, the project website contains information about safety measures when collecting data on-site and about data protection issues (e.g. data privacy, options to contact the project team in the case of questions, or to withdraw consent of use).
The citizenMorph website also contains information on project progress and results, displayed as a photo gallery, an event timeline, a list of publications, an embedded web map, and an external link to a dashboard that gives insights into the landforms reported by the citizens. The web map is the main element of the citizenMorph dashboard, which has several functionalities. For example, participants can search the map for the landform data they submitted (if a username was added). A pop-up appears with the    from the photo series using SfM techniques (figure 9). 3D models are generated only for those contributions where the photo series fulfilled basic requirements (several photos from different angles with a certain overlap). Before being included on the web map, citizens' contributions go through an automatic and expert-based quality check (to ensure consent of use granted, correct location, landform assignment, photo content).

Discussion
The combined use of participatory design, a prototyping model, and traditional requirement analysis methods revealed several elements required to build a citizen-centric system. Considering users' needs in project promotion, the design and implementation of instructions, information provision, feedback and recognition, and the design of the respective elements (figure 3; tables A1-A3), contributes to overcoming shortcomings in the quantity and quality of data reported by citizens.
Regarding project promotion, citizen representatives suggested various analogue and web-based techniques and tools. They highlighted the importance of in-person events, including the involvement of topic-related stakeholders to increase awareness of the project and the number of participants. The cit-izenMorph project took this into account through dedicated teaching activities and participation initiatives. These were carried out together with stakeholders, especially from the education sector. It must be highlighted that face-to-face contact is useful not only to increase interest, introduce the topic, explain the methods of data collection and reporting but also to build and maintain a project community as an important motivator for participants' ongoing engagement (McCully et al 2011, Cooper et al 2021). This is similar to other citizen science projects, such as Info Flora 1 or Goldschakal 2 , which include in-person events as part of their promotion strategy.
Citizen representatives identified several key aspects related to data contribution activities (table  A2), including instructions and tasks being clear, short, simple, and less work-intensive, which corresponds to findings in the literature (Rambaldi et al 2006, Newman et al 2010, Schaffer and Tham 2020. Considering citizens' needs is key in this context, as participation and survey design influence participation rates and data quality, i.e. validity, reliability, and consistency (Fagerholm et al 2021).
Before citizens start contributing data, they must be on-boarded to the citizenMorph system. Since not everyone who enters a website or application engages with it, becomes a user, or contributes data, this deserves special attention (Agarwal et al 2010, Morais et al 2013. In the citizenMorph system, the user on-boarding process embraces various aspects that address the different needs of the citizen representatives: (a) no registration and login process, (b) a brief introduction to clarify the importance of contributing to the project (text, video), and (c) contextrelated, clear, and brief instructions (text, audio files) with the possibility of accessing further information if desired (external links from the application). These aspects are in line with the need for a smooth user on-boarding process, considering that low access barriers at first contact with an application contribute to a higher number of participants. With cumbersome and inconvenient access, poor design, and lack of support, people will likely close an application without participating (Renz et al 2014, Fischer et al 2021. With regard to the user on-boarding process of the citizenMorph system, the topics of registration and login were discussed controversially during the test events. On the one hand, the possibility of contributing data without registration and anonymously was considered positive, simplifying and speeding up the contribution process and counteracting data privacy concerns. On the other hand, testers argued that having a registration option would enhance community building. Jay et al (2016) found that omitting the registration step increases contribution to citizen science projects by more than 60% but giving the option to create an account voluntarily raises contribution rates as well.
The use of a video to introduce the project (as part of the on-boarding process) was seen as a viable alternative to written text by the testers. This confirmed the questionnaire findings (table A2) and is also in line with other citizen science projects such as Picture Post 3 and Coastwards 4 , which use videos to present the topic, the aim, and the contribution process.
Citizens emphasised the key function of information provision in citizen science projects (table A2). Apart from project information and data contribution instructions, the questionnaire results (table A1) and prototype discussions with citizen representatives (figure 3) stressed the need to raise public awareness of citizen science and geomorphology. This is underlined by literature: Goudie and Viles (2014), for instance, found that the public's understanding and appreciation of geomorphology are lower than in other nature-related areas, like species and ecosystems. Thus, the question of how geomorphology can be made more relevant and engaging remains (Simm 2008), and as emphasized by Giusti (2010) the need for appropriate educational programs is clear.
The citizenMorph system addresses this in different ways, such as through in-person events and the use of different methods and media. Audio files (question-and-answer dialogue) to deliver data contribution instructions, branch-logic questions, and landform factsheets have been viewed positively by the testers in terms of inviting citizens to discover and discuss landforms together. Audio files are useful to foster collaboration (as preferred by the citizens; table A1) and increase the quality of the recorded landform data. While exploring landforms onsite, audio files are preferred, compared to texts, since intensive screen usage can be difficult outside, and users often dislike reading on the screen (Hennig and Vogler 2016). The information strategy used (figure 6) not only considers different needs to support differently skilled citizens, but also the fact that learning technology is usually related to a learning curve, so that after initial participation, citizens will need less support and, with more experience, contribute data related to a higher quality (Loiselle et al 2016, Cooper et al 2021. The citizenMorph system considered different ways in which participants receive feedback and recognition (figure 2). As in other citizen science initiatives (e.g. CrowdWater 5 , Did you feel it 6 ) geospatial tools, such as web maps, were used in the citizenMorph system. Web maps are particularly useful for citizen science projects as they not only provide feedback and recognition but are also useful to promote a project and attract participants because the public has increased interest and enjoys working with spatial data and related products (Craglia et al 2012, Thielmann et al 2012, Veenendaal et al 2017table A3).
Regarding the design, some of the testers, particularly younger ones, considered the citizenMorph design old-fashioned. They suggested considering the design of social media tools. This corresponds with Hennig and Vogler (2016), who highlight the different preferences of younger and older users. In the design of the citizenMorph system, we considered the demands of citizen representatives (figure 3), which generally matched recommendations regarding the usability and accessibility of websites (e.g. Usability.gov 2022; and met ethical requirements of participatory projects, such as attracting participants with different backgrounds and needs (Rambaldi et al 2006, Bonney et al 2009b. Other ethical concerns were also considered in the citizenMorph system, including issues described in the literature (Chambers 2006, Rambaldi et al 2006 that also corresponded with needs highlighted by citizen representatives, such as (a) no collection of personal data, (b) the possibility of withdrawing consent of use, (c) the provision of information (i.e. purpose, related benefits, stimulating learning), (d) providing feedback and recognition in a suitable way, and (e) using a participation and survey approach that is in line with citizens' reality (e.g. being considerate of people's time, not exposing people to danger, using technologies that can be mastered, non-repetitive activities).

Conclusions
Contributory citizen science projects and systems must be centred around citizen needs to overcome the challenges of data quantity and quality. Although the importance of these aspects for contributory citizen science is known, the question is how to consider these aspects from the citizens' point of view in the design and implementation of systems in ways that attract, inform, support, and motivate participation.
The results from the citizenMorph project, which informed the development of a pilot system to collect and report field data on landforms, emphasised the importance of in-person events for project promotion, suitable participation and survey design that is accompanied by a smooth user on-boarding process, increasing public knowledge on citizen science and the topic under consideration (i.e. geomorphology), giving feedback and recognition using elements that people enjoy using (e.g. geospatial tools), and a system design that is in line with usability and accessibility recommendations. Due to the different backgrounds of the citizens taking part in the development process, some features are controversial, such as whether to include a registration and login process as part of the citizens' on-boarding to the system, to which citizens relate benefits but also problems, and the general design of the system, which is viewed differently by different user groups (e.g. depending on their age).
The use of participatory design and the prototyping model fostered a citizen-centric development process, design, and implementation of the citizen-Morph pilot system. Based on our findings, we consider the applied approach to be an important step toward developing citizen science systems around citizen requirements because it allows in-depth identification and understanding of these aspects. However, even though the applied development approach helped reveal citizens' (previously unknown) requirements and showed how to comprehensibly take them into account when creating the citizenMorph pilot system, it also became obvious that collaborating with citizens requires significant additional effort. This includes selecting methods and tools suitable to be used by citizens and providing appropriate support for citizens to complete the tasks, for which a specifically composed team of scientists, developers, mediators, and particularly motivated citizens is key. The additional time, effort, and resources must be considered in advance of involving citizens in development processes.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).

Funding information
This research has been supported by the Austrian Science Fund (FWF) through the Top Citizen Science (TCS) project citizenMorph (FWF-TCS 47) and the project MORPH (FWF-P29461-N29) and the Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology of the Republic of Austria through the FFG Talente Praktikum program.

Ethical statement
The data collection approach (including measures to address ethical issues, such as the consent of use declaration) was reviewed and approved by the legal department/data protection coordinator of Salzburg University. The workshops were conducted in the context of high school and university events. The high school students involved in the project were older than 16 years.
The participants consented to have their data contributions being used in the study in the context of the described scientific purposes.