Building on the theoretical framing introduced above, VFEs can be understood as tools that scaffold multiple stages of the experiential learning cycle. In geoscience field education, learning often unfolds through direct observation in the field (concrete experience), followed by reflection, interpretation, and the development of conceptual models (reflective observation and abstract conceptualization), which are then tested through further field engagement (active experimentation).

Recent advances in digital geoscience, including digital outcrop models (DOMs), photospheres, photogrammetric reconstructions, and GIS-based visualization tools (e.g., Westoby et al., 2012; Howell et al., 2014; Horota et al., 2024), have expanded the ways in which these stages can be supported. VFEs allow students to preview field environments, revisit complex sites, and examine geological features at multiple scales. In this way, they may support anticipatory conceptualization before fieldwork, structured reflection after field experiences, and iterative reinterpretation across time and space (Fig. 1).

Rather than functioning as replacements for physical fieldwork, VFEs may extend experiential learning processes beyond the temporal and logistical constraints of in situ teaching. The extent to which students perceive these tools as supporting their experiential learning is explored in the Results and Discussion sections.

This research design was structured around the two research questions introduced in Sect. 1. Research Question 1 (RQ1), concerning students' perceptions of the usefulness of VFEs, was addressed through the Technology Acceptance Model (TAM) instrument and post-field educational impact items (Appendix A1). Research Question 2 (RQ2), concerning the management of novelty space, was examined primarily through pre-field preparedness items and post-field reflective responses (Appendix B1).

To address these questions, we adopted a mixed-methods approach to evaluate the integration of VFEs from the VR Svalbard platform into geoscience education at UNIS. The design focused on students' engagement with VFEs, their perceived educational value, and their reflections before and after physical fieldwork. Two key objectives guided the analysis: first, to evaluate technological readiness and perceived usability using TAM; and second, to assess students' perceived learning support and preparedness through structured survey instruments.

To investigate students' perceptions of digital field representations, a structured questionnaire combining Likert-scale items and open-ended responses was developed for this study. The study employed three interconnected survey components:

Technology Acceptance Model (TAM) survey. The TAM component assessed Perceived Usefulness (PU), Perceived Ease of Use (PEOU), and Behavioral Intention (BI) using items adapted from Davis (1989) and Venkatesh and Bala (2008), contextualized for digital field representations. Responses were recorded on a 7-point Likert-type scale (1 = extremely unlikely; 7 = extremely likely).

The survey components were distributed at two distinct time points. The pre-field instrument was administered after students had engaged with VFEs but prior to the physical field excursion. The post-field instrument was distributed following completion of field activities. The TAM component was administered after students had sufficient exposure to the VR Svalbard platform to evaluate its usability and usefulness.

Together, these instruments enabled examination of technological acceptance (RQ1), perceived educational support (RQ1), and perceived management of novelty space (RQ2), while allowing interpretation through the lens of Experiential Learning Theory.

A total of 131 students enrolled in relevant UNIS courses were invited to participate. Sixty-nine students consented to participation, and 66 completed the main survey components. Of these:

54 respondents completed the TAM section,

The study involved 131 students attending bachelor's, master's, and PhD-level courses at UNIS (Table 2). Participants were selected based on their participation in field-based courses at UNIS that incorporated VFEs into their curriculum. These courses span a range of topics, ensuring a sample representative of various subfields within geoscience and polar meteorology (Table 2). Participants provided informed consent and were briefed on the study's objectives and data collection processes.

The selection of UNIS courses allowed the study to target students with significant exposure to Arctic fieldwork, offering a unique context in which to evaluate the integration of VFEs into traditional field education.

The survey data reveal significant engagement with the VR Svalbard platform among participants. Of the 66 students who completed the survey, 86 % reported having used the platform at least once. The use of VFEs was integrated into several contexts, demonstrating its broad applicability within field-based education. Notably, 71 % of respondents used the VFEs during field preparation lectures, while 55 % used them as part of post-fieldwork exercises. Additionally, 53 % of students engaged with the platform independently out of personal curiosity, showing a high degree of self-motivated exploration (Fig. 4).

Beyond course-integrated use, students also employed the tool for planning their own fieldwork (38 %), revisiting specific field localities (53 %), and extracting visual data for assignments or term projects (38 %). This range of uses emphasizes the platform's flexibility in serving various pedagogical and research needs.

In terms of time spent, 42 % of users reported engaging with the platform for less than one hour, 40 % for one to three hours, 15 % for three to five hours, and 4 % for over five hours. These time ranges likely reflect the one-hour session typically allocated in class for VR Svalbard activities. However, it is especially encouraging to observe that more than half of students used the platform beyond the scheduled class time. This suggests that many students engaged with the platform beyond scheduled sessions, suggesting its effectiveness not only as a classroom tool but also as a meaningful extension of field learning. These engagement patterns indicate that VFEs were used both as structured instructional tools and as optional self-directed resources.

The technical aspects of VR  albard's VFEs were evaluated using a TAM questionnaire. Respondents were asked to rate the likelihood represented by each statement on a 7-point scale. The dataset included responses from 54 participants who completed this section of the questionnaire.

Students rated the platform highly in terms of perceived usefulness. The highest-scoring items were “Using VR Svalbard would enhance my effectiveness in the field” and “I would find VR Svalbard useful in my field learning activity,” each with a mean score of 6.0. These high ratings underscore students' belief that the platform positively contributes to their learning outcomes. Other positively rated items included the platform's ability to help complete field tasks more quickly and efficiently.

Perceived Ease of Use statements also scored well, with statements such as “Learning to operate VR Svalbard would be easy for me” and “My interaction with VR Svalbard would be clear and understandable” both receiving mean scores around 5.9 (Fig. 5). Although students generally found the platform intuitive, slightly lower ratings were given for items related to flexibility and user control, suggesting opportunities for improvement in the interface and customization features.

Behavioral Intention responses indicated that students were likely to continue using the platform. Statements such as “I would recommend VR Svalbard to my peers” and “I would use VR Svalbard for future field activities” received high scores, reflecting a strong willingness among students to engage with the tool beyond a single course context (Fig. 5). This finding is particularly significant, as it suggests the long-term adoption of VR Svalbard as a resource for students. Moreover, it indicates that students are not only receptive to technology but may also serve as advocates for its broader implementation within geoscience curricula.

Together, the high ratings across Perceived Usefulness, Perceived Ease of Use, and Behavioral Intention suggest that VR Svalbard meets key criteria for technology acceptance in educational settings. The platform was perceived as a useful, easy-to-learn tool that students intend to continue using, an important foundation for promoting digital literacy and innovation in field-based geoscience education. However, the slightly lower ratings related to control and flexibility indicate areas for design improvement, particularly in refining user experience to better support a variety of learning preferences and workflows.

To interpret how VFEs supported experiential learning processes, Likert-scale responses were grouped according to their alignment with stages of Kolb's experiential learning cycle (Kolb, 1984) and examined alongside the concept of novelty space.

This section presents an analysis of how students perceived the educational value of VFEs in relation to their geoscience learning. Drawing on quantitative Likert-scale responses and qualitative open-ended comments from Appendix B1, the findings highlight how the integration of VFEs before and after field activities influenced student engagement, spatial awareness, immersion, preparedness, and perceived learning outcomes (Fig. 6). These themes are further substantiated by the full distribution of responses presented in Fig. 6, which provides a more nuanced view of how students positioned themselves. This scale allowed students to express varying degrees of agreement with statements about their experiences with the VFEs.

VR Svalbard was introduced into courses through a structured, two-phase integration strategy: pre-field and post-field activities. This deliberate inclusion aimed to optimize learning outcomes by scaffolding students' engagement with geoscientific content.

In the pre-field phase, instructors used VR Svalbard to orient students to Arctic landscapes and specific field locations. Activities included navigation exercises, spatial reasoning tasks, and exposure to realistic terrain models. These sessions helped students build mental models of field environments, thus increasing their confidence and readiness before the actual fieldwork.

In the post-field phase, VR Svalbard VFE served as a reflective tool. Students revisited sites virtually to reinforce learning, compare virtual models with real-world observations, and identify aspects they may have missed in the field. This allowed for deeper engagement with the material and supported students' ability to connect observations across scales and contexts.

Open-ended responses highlighted areas for improvement, including navigation refinements, faster loading speeds, and additional guided exercises. Students also expressed interest in expanded use across courses, with 84 % indicating willingness to use VR Svalbard for future field preparation.

Selected student comments illustrate both strengths and limitations:

“Seeing the site in summer helped me understand what was under the snow.”

Developing the VFEs at UNIS was a gradual process that builds on incremental improvements and systematic data acquisition. We have been acquiring digital outcrop models at UNIS since 2017 and openly sharing them through the Svalbox portal (Betlem et al., 2023). We started acquiring aerial photospheres and building the VRSvalbard platform in 2021 (Horota et al., 2024). The online platforms themselves are affordable (approximately 10 000 EUR in total for development and ca 1000 EUR per year for maintenance costs – web hosting largely) and sustainable for long-term maintenance. Data acquisition, however, is costly. A usual single field day in the high Arctic can easily cost 1000 EUR covering transport in open boats or snowmobiles. UAVs used at UNIS (mostly DJI Mavic 2 Pro and Mavic 3) have exponentially improved in quality over the past decade and have modest costs (ca. 2000 EUR per UAV including controller). Education at UNIS is fully financed by the Norwegian Ministry of Education and Research. On average, educating one student-year (i.e. 60 ECTS) at UNIS costs ca. 10 000 EUR. In practice, this means that the courses studied by us (Table 2; typically 10–15 ECTS courses for up to 20 students) operate on total budgets of about 30 000–50 000 EUR including all field costs and travel/salary for guest lecturers, but excluding salary for UNIS-based course responsible staff. Field costs vary greatly between courses depending on the access to field site (walking, boating or snowmobiling), whether courses include overnight stay (cabins or hotel ships) and length of field work. With such high costs it is imperative that digital data (photospheres and digital outcrop models) are acquired opportunistically during courses or as part of research-based fieldwork.

In summary, running VFEs in general requires moderate resources and should be easily adoptable at most geoscience institutions. At UNIS, where the field access costs are significantly higher, VFEs are a must.

Before evaluating pedagogical impact, it was necessary to establish whether the platform itself was accepted as usable and relevant. High scores across Perceived Usefulness, Perceived Ease of Use, and Behavioral Intention indicate that students viewed VR Svalbard as intuitive and educationally meaningful. This finding is important because technological friction can undermine learning outcomes (Davis, 1989; Venkatesh and Bala, 2008). If cognitive resources are diverted toward navigating an interface, they cannot be fully allocated to disciplinary reasoning. The strong acceptance observed here suggests that students' reported gains in preparedness and reflection were not confounded by usability barriers. Moreover, these results challenge assumptions that students are inherently comfortable with educational technology. Acceptance must be empirically demonstrated rather than presumed. In this case, technological usability appears to have enabled rather than obstructed experiential learning processes.

Interpreted through Experiential Learning Theory (Kolb, 1984), the findings suggest that VFEs supported multiple stages of the learning cycle. Pre-field engagement appeared to support Abstract Conceptualization by enabling students to construct preliminary mental models before entering the field. Students reported improved geographic orientation and familiarity, indicating that virtual previews contributed to anticipatory structuring of knowledge. Physical fieldwork then provided Concrete Experience, where sensory, spatial, and environmental dimensions of learning unfolded in situ. Post-field revisiting aligned with Reflective Observation. Students reported that returning to digital representations helped consolidate understanding and reinterpret previously observed features. In this way, VFEs extended reflection beyond the temporal constraints of field excursions. Independent exploration and optional engagement outside scheduled activities may reflect Active Experimentation, where learners test interpretations and navigate field sites in self-directed ways. Rather than compressing experiential learning into a limited field window, VFEs redistributed it across time and contexts. This temporal extension may be particularly valuable in environments where logistical constraints restrict repeated site visits.

Field-based education inherently involves novelty. In remote Arctic environments, students must process unfamiliar terrain, safety procedures, weather conditions, and spatial complexity simultaneously. Novelty space describes the cognitive and affective load associated with these first encounters (Stainfield et al., 2000). Students in this study reported increased preparedness and orientation following virtual previews. While the study did not directly measure cognitive load, the consistent pattern of improved familiarity suggests that VFEs may moderate environmental novelty prior to physical exposure. Importantly, managing novelty does not imply eliminating it. Some degree of unfamiliarity is pedagogically valuable and contributes to disciplinary identity formation. However, reducing excessive environmental disorientation may allow students to allocate greater cognitive resources toward geological interpretation rather than situational adjustment. In this sense, VFEs may calibrate novelty space rather than remove it.

Students did not strongly support replacing physical fieldwork with virtual alternatives. This reinforces existing studies emphasizing that VFEs are most effective when integrated alongside, rather than instead of, field experiences (Cliffe, 2017; Pugsley et al., 2022).

The findings suggest that students perceive digital field representations as:

Tools for preparation

The accessibility implications of VFEs may also extend beyond physical access to field sites. For some students, including neurodivergent students or those with limited prior field experience, unfamiliar field environments, social situations, safety procedures, and logistical uncertainty may increase anxiety and cognitive load before field activities begin. In this sense, VFEs may function partly like preparatory “social stories” (Gray and Garand, 1993), allowing students to preview what a field site, route, or activity may involve before encountering it in person. However, virtual environments may also constitute a novelty space in themselves, particularly for students with limited exposure to digital or immersive tools. Given that 42 % of users engaged with VR Svalbard for less than one hour, future research should examine how technological novelty interacts with field novelty and how both influence learning outcomes.

In addition, VFEs provide access to localities that may be seasonally restricted, weather-limited, or logistically inaccessible. In the context of environmental sustainability and increasing scrutiny of travel-related carbon footprints, digital extensions of field learning may also offer opportunities to balance experiential depth with ecological responsibility.

Several limitations should be acknowledged. First, the study relies on self-reported perceptions rather than direct measurement of learning gains. While perceived preparedness and orientation are important indicators, future research should examine measurable impacts on spatial reasoning and field performance. Second, the study was conducted within a single institutional context in the High Arctic. The extent to which findings generalize to other geoscience settings requires further investigation. Third, novelty space was inferred from preparedness and orientation measures rather than directly operationalized through cognitive load instruments. Future studies could incorporate validated scales to assess cognitive and affective novelty more explicitly. Future research should also explore co-design approaches involving students and instructors to refine interactivity, task guidance, and integration strategies within VFEs.