Introduction

Despite mounting evidence of climate change, public engagement remains limited, partly due to psychological distance. Addressing this gap requires not only interventions that promote climate action but also an understanding of the psychological mechanisms that shape climate-related behaviors. For instance, a large mega-study revealed that upon comparing various behavioral interventions for promoting climate beliefs, interventions that focus on decreasing psychological distance were more effective than those such as negative emotional induction, future-self continuity, dynamic social norms and 8 other types of behavioral interventions1.

One of the primary psychological barriers to climate engagement is psychological distance, which is grounded in two key theoretical frameworks. The first is the discounting principle, which suggests that people perceive future consequences as less valuable than immediate ones. As a result, individuals tend to discount the risks of climate change, including rising sea levels for coastal areas, perceiving them as distant in both time and space, thereby diminishing their sense of urgency and action2,3,4. The second framework, construal level theory, proposes that people mentally represent events and objects in either concrete or abstract terms, which in turn influences their decision-making and behavior. In this vein, psychological distance has been explicated into four dimensions: (i) hypothetical distance, wherein individuals perceived certainty with an event is associated with a future event, (ii) temporal distance, wherein individuals regard an event to happen in a distant future, (iii) spatial distance, wherein individuals perceive the impacts of an event happening in a geographically distant area, and (iv) social distance, where one finds that while these events happen, they would not happen to those like themselves5. Research on spatial psychological distance has illustrated that those who have more recent and vivid direct experiences with climate change and extreme weather events are more likely to express greater concern and support for environmental action6,7. This is because psychological distance is associated with cognitive abstraction, wherein greater psychological distance is related to higher cognitive abstraction, and lower concreteness5,8. Similarly, psychological distance is related to personal relevance, wherein increasing personal relevance to climate events by decreasing psychological distance, is related to increased concern for climate events9. However, for individuals to act on climate change’s consequences, such as supporting pro-environmental charities and policies, there is a need to traverse this psychological distance.

Place attachment refers to the emotional bond that develops between individuals and specific geographical locations, encompassing the affective, cognitive and behavioral relationship that individuals have with the environment around them33. While traditional conceptualizations of place attachment treat it as a relatively stable trait construct that develops over extended periods through repeated experiences and meaningful interactions with a location51, recent work has suggested that place attachment may also exhibit state-like properties, capable of being temporarily activated through immersive experiences that provide contextual cues33. This psychological construct is particularly relevant to environmental psychologists because research demonstrates that individuals are more likely to support environmental protection and engage in pro-environmental behaviors for places to which they feel emotionally connected49.

The way climate change is communicated further influences psychological distance and public engagement. People respond more to direct experiences than to abstract statistics2. Research has illustrated that statistical descriptions of climate risks fail to elicit action because climate change is a slow-moving and often invisible process that cannot be experienced directly10. However, surveys have illustrated that personal experiences with extreme weather events can influence risk perceptions and policy support11,12. In addition, psychological distance is related to discrete emotions, wherein increasing distance decreases emotional intensity and affects climate change engagement13. Nevertheless, effectively communicating the urgency of climate change requires more than proximizing the issue as empirical evidence has shown mixed results for psychological distance-related interventions, particularly as climate change effects become widespread14.

One critical moderating factor in climate communication research, particularly in the United States where climate change has become highly politicized along partisan lines is political orientation69. Prior work has illustrated that conservative identifying individuals have strongly lower climate change beliefs than those identifying as liberal70. This ideological divide extends beyond beliefs to receptiveness to climate communication interventions with research showing that the extent of different interventions appealing to conservatives and liberals40. Thus, understanding how political orientation moderates responses to climate communication interventions is crucial for developing effective messaging strategies, as interventions that work for one political group may be ineffective or even counterproductive for another.

Consequently, the medium through which climate information is communicated plays a crucial role in shaping public perceptions, emotions, and behaviors. Studies comparing different media formats suggest that immersive experiences can help reduce psychological distance and enhance climate risk perceptions. For example, research indicates that viewing climate change events in virtual reality (VR) is more effective than watching them on a 2D screen15. Similarly, compared to traditional videos and magazine articles, VR has been found to evoke stronger negative emotions, decrease meat and dairy consumption intentions, and enhance feelings of awe and spatial presence16. Beyond environmental behaviors, VR has also been shown to promote prosocial behavior and empathy. Notably, performing actions in VR, such as flying, has been linked to greater prosocial outcomes in participants17,18,19.

VR’s effectiveness in climate communication can be attributed to two key psychological affordances: presence and plausibility. Presence refers to the extent to which individuals feel physically immersed in a virtual environment20, while plausibility describes the perceptual illusion that what is happening in the virtual world is real, despite the user knowing otherwise21. These features allow participants to experience climate events in a way that feels immediate and tangible, eliciting stronger emotional and cognitive responses. Presence has been found to significantly heighten emotional arousal22 and the perceptions of negative emotions23. Due to the tight association between presence and plausibility, research has also illustrated that the plausibility illusion enhances emotional arousal and perceived fear when encountering fearful events in VR24.

While emotional elicitation has been widely studied in VR25,26, research on VR and eco-emotions, or emotions that are particularly related to the climate crisis and the environment, remains scarce. Eco-emotions are classified into two categories: (i) adaptive emotions, which promote greater engagement in pro-climate and environmental behaviors such as eco-anger, frustration, and climate hope, (ii) and non-adaptive emotions, such as eco-anxiety, eco-depression, and eco-sadness, which result in lower climate engagement and wellbeing, thereby increasing worry and misery27,28,29. Additionally, the spatial remoteness of climate change impacts often leads to indifference and disregard for the suffering of those affected and thereby also leads to greater inaction5,30.

Building on this literature, our work explores the potential of VR to bridge psychological distance and elicit adaptive eco-emotions more effectively than static images by situating individuals in both proximal and distant places. Participants fly in a geographic area while listening to an audio recording of a climate change-related news event occurring in the very location they are virtually visiting. Our goal is to examine whether situating individuals in distant places via VR can make those places feel psychologically closer to a foreign place, through mechanisms such as the place illusion and plausibility illusion, and, in doing so, promote climate engagement. We posit that this effect is driven by situated perception, whereby individuals extract emotional and cognitive meaning from the environment they are immersed in, using that contextual information to evaluate the significance of what they hear and experience31. Importantly, although constructs like place attachment are often treated as long-term and stable, prior research on situated learning suggests that immersion in foreign contexts can serve as powerful learning and emotional experiences, even in brief encounters32. Such immersive exposure can provide contextual cues that help individuals make sense of complex information, foster understanding from new perspectives, and even stimulate emotional connections to unfamiliar environments. This opens the possibility that place attachment may also function as a state-level response under certain immersive, context-rich conditions33.

Unlike prior VR interventions that create new experiences, we leverage an existing online geographic VR platform, FlyVR, to immerse participants in distant locations. By allowing individuals to virtually fly over these locations while listening to climate-related news, we test whether VR’s presence and plausibility illusions can make remote climate events feel more immediate.

To this end, we pose the following research questions:

(1) Does VR reduce psychological distance and enhance climate engagement for distant locations compared to static images?

(2) Does VR elicit stronger adaptive eco-emotions, such as climate hope and eco-anger, while minimizing non-adaptive eco-emotions, such as eco-anxiety and eco-depression?

(3) How does political orientation moderate the relationship between VR exposure and climate engagement?

If this prediction holds true, VR could be a promising tool for behavioral interventions that bring climate change events from distant spatial locations to our present reality and stimulate action.

Results

Media and spatial boundary on psychological distance

Findings illustrated a significant main effect of VR (b = −43.22, SE = 6.74, t (146) = −6.41, p < 0.001) (Table 1), indicating that participants in the VR condition (M = 85.57, SD = 24.84) scored significantly lower on our psychological distance measure for cognitive abstractedness compared to viewing static images (M = 85.57, SD = 24.84). While no main effect of spatial boundary (b = −11.27, SE = 7.05, t (145) = −1.60, p = 0.11) was found, significant interaction effects between the media conditions and spatial boundaries were elicited (b = 21.99, SE = 10.15, t (143) = 2.17, p = 0.03). Participants who were in static images viewing a distal condition (M = 123.42, SD = 32.16) scored significantly higher than both VR in a distal condition (M = 81.93, SD = 21.89) and VR in a proximal condition (M = 90.28, SD = 27.83). Participants who were in static images and proximal condition (M = 113.06, SD = 36.92) also scored significantly higher than those in the VR distal condition and VR local condition (Fig. 1). No significant differences were found between the static images condition irrespective of spatial boundary and between the VR conditions and spatial boundary. Political orientation was not a moderator of this manipulation check (b = 0.18, SE = 0.13, t (125.38) = 1.40, 95% CI [−0.07, 0.43], p = 0.16).

Fig. 1
figure 1

Psychological distance levels by media and spatial boundary conditions.

Full size image
Table 1 Fixed effects of VR on psychological distance and dependent variables.
Full size table

Media and spatial boundary conditions on climate emotions

Findings illustrated a significant main effect of VR (b = −2.00, SE = 0.69, t (116) = −2.91, p < 0.01, 95% CI [−0.02, 0.88]]) on climate indifference wherein those in the VR condition felt less indifference (M = 4.28, SD = 2.90) than those viewing static images (M = 4.47, SD = 2.96) (Table 1). There was also a main effect of spatial boundary (b = −2.02, SE = 0.66, t (116) = −3.37, 95% CI [−0.29, 0.61], p < 0.01) wherein those in the distal condition (M = 4.64, SD = 3.19) had higher climate indifference than those in the proximal condition (M = 4.12, SD = 2.63). There were significant interaction effects between VR and spatial boundary (b = 2.90, SE = 0.98, t (116) = 2.95, 95% CI [−1.34, −0.04], p < 0.01) wherein those in the VR and distal condition (M = 3.70, SD = 3.14) had less climate indifference than those in the no VR distal condition (M = 5.33, SD = 3.09) (Fig. 2). Participants viewing static images in the distal location (M = 5.33, SD = 3.09) had significantly higher climate indifference than those in the proximal location (M = 3.48, SD = 2.51). Political orientation moderates these effects (b = 0.03, SE = 0.01, t (116) = 3.00, 95% CI [−1.34, −0.04], p < 0.01), the more conservative participants self-reported themselves to be compared to those reporting more liberal orientation, the less likely that the media and spatial boundaries reduced climate indifference (Fig. 3).

Fig. 2
figure 2

Climate indifference (top-left), climate frustration (top-right), place attachment (bottom left) and awe (bottom right) levels by media and spatial boundary conditions.

Full size image
Fig. 3
figure 3

Political orientation as a moderator for various climate emotions and climate risk perceptions.

Full size image

We found similar main effects of VR on climate frustration (b = 1.31, SE = 0.47, 95% CI [0.17, 0.99], t (133) = 2.81, p < 0.01) such that those in the VR condition have a higher degree of climate frustration (M = 7.54, SD = 1.99) than those viewing static images (M = 7.14, SD = 2.47) (Table 1). There was also a significant main effect of spatial boundary on climate frustration (b = 1.54, SE = 0.46, 95% CI [0.28, 1.09], t (133) = 3.33, p < 0.01), such that those in the proximal condition (M = 7.55, SD = 2.00) had a higher degree of frustration than those in the distal condition (M = 7.13, SD = 2.47). Political orientation moderates these effects (b = −0.05, SE = 0.10, 95% CI [−0.57, −0.28], t (133) = −5.71, p < 0.001), wherein participants who self-reported themselves to be more conservative expressed less climate frustration following the intervention compared to those who self-reported themselves to be more liberal (Fig. 3). There was also a significant interaction effect between VR and spatial boundary (b = −1.62, SE = 0.67, 95% CI [−1.31, −0.13], t (133) = −2.43, p < 0.05), wherein those in the VR distal condition (M = 7.66, SD = 2.20) had a higher degree of climate frustration than those in the static images and distal condition (M = 6.62, SD = 2.63) (Fig. 2). Participants viewing static images in the proximal location (M = 7.66, SD = 2.22) had significantly higher climate frustration than those in the distal location (M = 6.68, SD = 2.69).

Similarly, there was a significant main effect of VR (b = 1.11, SE = 0.52, 95% CI [0.03, 0.91], t (130.35) = 2.13, p < 0.05) on climate helplessness wherein those in the VR condition felt more helplessness (M = 7.29, SD = 2.31) than those viewing static images (M = 6.64, SD = 2.43) (Table 1). There was also a significant main effect of spatial boundary (b = 1.10, SE = 0.52, 95% CI [0.03, 0.90], t (130.50) = 2.11, p < 0.05) wherein those in the proximal condition (M = 7.30, SD = 2.13) felt higher climate helplessness than those in the distal condition (M = 6.62, SD = 2.58). Political orientation moderates these effects (b = −0.04, SE = 0.01, 95% CI [−0.50, −0.19], t (126.15) = −4.34, p < 0.001), wherein those who self-reported themselves to be more conservative felt less climate helplessness than those who self-reported themselves to be more liberal following the intervention (Fig. 3). There were no interaction effects between VR and spatial boundary on climate helplessness. No significant differences in climate helplessness were found between those who viewed a static image in the distal or proximal location.

There was a significant effect of spatial boundary on climate anger (b = 1.40, SE = 0.55, 95% CI [0.12, 0.96], t (94.83) = 2.55, p < 0.05) (Table 1). Political orientation moderates these effects (b = −0.04, SE = 0.01, 95% CI [−0.47, −0.19], t (36.91) = −4.66, p < 0.001), with individuals identifying as more conservative to be less angry than those identifying as more liberal. However, no significant main effect of VR on climate anger but significant interaction effects of VR and spatial boundary were found (b = −1.64, SE = 0.80, 95% CI [−1.25, −0.02], t (117.72) = −2.04, p < 0.05). Participants viewing static images in the proximal location (M = 7.23, SD = 2.18) had marginally higher climate anger than those in the distal location (M = 6.45, SD = 2.73).

No significant main or interactions effects of VR or spatial boundary on climate depression, hope, misery, anxiety, fear, or guilt were found. Political orientation significantly moderated the effect for depression (b =−0.04, SE = 0.01, 95% CI [−0.50, −0.18], t (127) = −4.02, p < 0.001), hope (b = 0.03, SE = 0.01, 95% CI [0.04, 0.39], t (122) = 2.58, p < 0.05), misery (b = −0.05, SE = 0.11, 95% CI [−0.55, −0.23], t (130) = −4.96, p < 0.001), anxiety (b = −0.04, SE = 0.01, 95% CI [−0.51, −0.19], t (135) = −4.28, p < 0.001), and fear (b = −0.04, SE = 0.01, 95% CI [−0.55, −0.23], t (133) = −4.88, p < 0.001) (Fig. 3). Overall, participants who identify as more conservative feel less depressed, miserable, anxious, fearful about the climate and more hopeful than those identifying as more liberal. No significant effects of VR, spatial boundary or political orientation on climate guilt were found.

Media and spatial boundary conditions on place attachment

Results illustrated that VR had a significant main effect on place attachment (b = 0.70, SE = 0.23, 95% CI [0.30, 1.22], t(127) = 3.05, p < 0.01) wherein VR created more place attachment (M = 1.65, SD = 0.28) compared to static images (M = 1.47, SD = 0.36) (Fig. 2; Table 1). No significant differences in place attachment were found between those who viewed a static image in the distal or proximal location. Political orientation did not moderate the effects.

Media and spatial boundary conditions on climate storytelling length and language features

Results found significant effects of VR on climate storytelling investment, i.e., the length of stories participants were asked to write following the experimental conditions (b = 3.54, SE = 0.60, 95% CI [0.76, 1.60], t(133) = 5.95, p < 0.001) (Table 1). Moreover, there was a significant interaction effect between VR and spatial boundary (b = −2.02, SE = 0.86, 95% CI [−0.19, 0.11], t(133) = −2.34, p < 0.05) such that participants were more invested in climate storytelling following the VR experience when in the distal VR condition (M = 16.15, SD = 3.61) compared to viewing a static image of a distant location while listening to a climate change news story of that corresponding place (M = 12.38, SD = 2.96). No significant differences in climate storytelling length were found between those who viewed a static image in the distal or proximal location. However, there were no significant main effects of spatial boundary and political orientation on climate storytelling investment.

On measuring the amount of negative emotional words in each story, we found no significant main or interaction effects of VR and spatial boundary respectively. Political orientation was a moderator on negative emotion word frequency wherein those with more conservative beliefs had lower negative emotions (b = −0.004, SE = 0.002, 95% CI [−0.34, −0.01], t (136) = −2.35, p = 0.02).

Media and spatial boundary conditions on awe

Results found that VR had a significant main effect on awe (M = 1.61, SD = 0.25) compared to static images (M = 1.44, SD = 0.32) (b = −0.22, SE = 0.06, 95% CI [−0.67, 0.59], t (136) = −3.39, p < 0.01) (Fig. 2; Table 1). No significant main and interaction effects with spatial boundaries were found. Political orientation did not moderate the effects.

Media and spatial boundary conditions on climate risk perceptions

Results found that there was a significant main effect of VR on climate risk perceptions (b = 0.28, SE = 0.11, 95% CI [0.04, 0.84], t (136) = 2.49, p < 0.05) wherein those experiencing a climate change news story in VR reported higher risk perceptions (M = 5.05, SD = 3.16) than those viewing static images (M = 5.03, SD = 0.59) (Table 1). There was a significant main effect of spatial boundary on climate risk perceptions (b = 0.28, SE = 0.11, 95% CI [0.03, 0.82], t(136) = 2.43, p < 0.05) wherein those in the proximal condition (M = 5.05, SD = 0.56) had higher risk perceptions than those in the distal condition (M = 5.02, SD = 0.61). No significant differences for climate risk perceptions were found between those who viewed a static image in the distal or proximal location. There were significant interaction effects between VR and spatial boundary (b = −0.45, SE = 0.16, 95% CI [−1.30, −0.16], t (136) = −2.78, p < 0.01), with post-hoc tests illustrating marginal effects for increased risk perception in VR for distal locations (M = 5.13, SD = 0.52) compared to static images (M = 4.94, SD = 0.63) (p = 0.07). Political orientation moderates these effects (b = −0.14, SE = 0.02, 95% CI [−0.65, −0.37], t (136) = −7.38, p < 0.001), that is, the more conservative an individual self-reported themselves to be the less they reported climate risk perceptions (Fig. 3).

Media and spatial boundary conditions on spatial presence

Results showed that VR has a significant main effect on spatial presence (b = 0.06, SE = 0.24, 95% CI [0.17, 1.06], t (136) = 2.33, p < 0.01) wherein VR creates more spatial presence (M = 1.66, SD = 0.26) compared to 2D images (M = 1.45, SD = 0.40) (Table 1). Those who experienced a climate change news story in VR in a proximal location had significantly higher spatial presence (M = 1.66, SD = 0.24) than those who did not experience the story in VR (M = 1.45, SD = 0.42) (b = −0.23, SE = 0.08, 95% CI [−0.60, 0.68], t(135) = −2.77, p < 0.05). No significant differences were found in spatial presence between those who viewed a static image in the distal or proximal location.

Discussion

Using a between-subjects experiment, we found that experiencing a climate story about a distant location in VR significantly decreased climate indifference and increased climate frustration compared to both static images and VR in a proximal location. When analyzing responses to static images alone, individuals exhibited lower concern for climate events in distant locations than in proximal ones. Given prior research on the differential effects of climate emotions, where frustration is considered an eco-adaptive emotion, while emotions such as misery, anxiety, fear, and guilt are less adaptive, our findings suggest that VR elicits a more constructive emotional response to climate change compared to static imagery, potentially fostering greater engagement with climate issues27. This is consistent with the debate on psychological distance and construal level theory which states that an abstract construal level causes people to act more in line with their abstract personal values, rather than responding to situational concerns14,63,64.

We also show that VR increases place attachment for distant locations compared to static images. Given that prior work has found that one’s scale of environmental action is connected to a sense of place, our findings illustrate VR’s potential in connecting individuals to distant places49,50. While place attachment is often conceptualized as a relatively stable trait construct that develops over time51, recent research suggests that it can also exhibit state-like properties33. Situated perception provides a compelling lens to understand how immersive environments like VR can temporarily evoke a sense of “being in” a place, which then allows individuals to interpret emotional and informational cues more deeply31. This is consistent with research on situated learning, which shows that immersion in foreign or novel environments can prompt learners to form emotional and conceptual connections through exposure to contextualized cues32. In this way, VR simulates the environmental and sensory richness necessary to foster state-level place attachment, even for distant or unfamiliar locations. Moreover, the congruence between being in the place (even virtually) and receiving information about it may enhance emotional salience, making the place feel personally relevant. This supports the notion that VR can momentarily “activate” place attachment, providing a foundation for more enduring connections if exposure is sustained over time49,50.

Additionally, experiencing a climate change news story in VR marginally increased climate risk perceptions for distant locations compared to viewing static imagery. However, this effect is not present for proximal locations. This aligns with the argument that VR is particularly effective for experiences that are otherwise dangerous, impossible, counterproductive, or expensive52. Since many climate-related impacts occur in distant locations, VR can help bridge this psychological gap by enhancing place attachment and increasing risk perceptions, particularly for remote locations26.

VR also heightened engagement with climate storytelling, as measured by the number of words participants wrote in their climate narratives. Political orientation did not significantly affect storytelling investment, with conservatives and liberals demonstrating similar levels of engagement across both VR and static image conditions. This finding contrasts with the emotional and risk perception measures, where political orientation served as a significant moderator, suggesting that behavioral measures of storytelling engagement may be less susceptible to political polarization than self-reported attitudes and emotions53.

Consistent with prior studies, we found that VR elicited positive emotions such as place attachment, awe, and spatial presence, compared to listening to a news story while viewing static images54,55. These effects were not moderated by political orientation. However, given that political affiliation is a key predictor of climate policy support and skepticism56,57, we note that interventions focused on decreasing psychological distance, as in our study, tended to enhance climate emotions and risk perceptions more strongly among liberals than conservatives. This finding is consistent with prior research on the effectiveness of psychological distance interventions in the context of ideological divides40.

Altogether, these findings highlight VR’s efficacy in fostering place attachment and reducing psychological distance for distant locations, contributing to the growing body of evidence on how immersive experiences can shape climate engagement and can be understood through the lens of the two key psychological mechanisms we outlined: presence and plausibility illusions. As VR promotes the feeling of presence, which allows them to be physically immersed in the climate-affected location and plausibility, which allows users to perceive virtual experiences as real despite knowing it was simulated, we can better understand why VR was particularly effective for distant locations, where participants typically lack direct sensory experience20,21. The presence illusion allowed participants to feel as though they were visiting remote climate-affected areas, while the plausibility illusion made the climate impacts feel immediate and tangible rather than abstract24.

This work integrates the role of emotions in psychological distance research, demonstrating that situating individuals virtually within a climate news story facilitates more concrete mental representations of place compared to viewing static images. This, in turn, fosters adaptive emotions that enhance climate risk perceptions13.

However, future research is needed to disentangle whether VR exploration primarily promotes familiarity with an area, thereby increasing concern for it, or whether VR fosters identification with the people in impacted places by simultaneously presenting a climate event occurring in the visited location13,42. Future studies should explore the variety of VR experiences that might encourage positive climate action58,59. While our study focused solely on how exposure to a news story about climate change-induced flooding affected climate emotions and risk perceptions, it would be valuable to investigate whether narratives about overcoming climate-related disasters can foster greater community resilience and inspire climate action by evoking feelings of hope and awe29,60.

Another important limitation of our study is that both proximal and distant locations were constrained within US boundaries. This design choice was made to control cultural, linguistic, and political factors that might confound the effects of spatial distance alone. However, this limits the generalizability of our findings to truly international contexts. Moreover, while this study collected participants’ racial and ethnic identification, we did not measure their sense of American identity or attachment to the United States. This represents a possible limitation, as participants with stronger national identification might perceive all US locations as relatively proximal regardless of geographic distance, potentially attenuating our spatial boundary effects67,68.

Future research should examine whether VR interventions show similar efficacy when bridging psychological distance to climate-affected locations in other countries, where cultural differences65, language barriers, and different climate vulnerabilities66 might moderate the effects we observed. Given that many of the most severe climate impacts occur in developing nations that may feel culturally and politically distant to participants in developed countries64, understanding VR’s effectiveness across international boundaries represents a critical area for future investigation.

Moreover, our findings should be interpreted in the context of recent theoretical debates about the construct of psychological distance as a recent comprehensive review suggests that the role of psychological distance in climate change engagement may be overestimated, arguing that people do not actually perceive climate change as psychologically distant as previously assumed14. Yet, psychological distance interventions have been seen to be useful in a prior largescale showing that among 11 expert-crowdsourced interventions, psychological distance reduction was the most effective in strengthening climate change beliefs1. These results, along with the current results, suggest the psychological distance interventions may have value in particular contexts, the interaction effects observed in this study, for instance, where VR was particularly effective for distant, rather than proximal locations, align with the idea that psychological distance may matter more for some climate impacts than others. Additionally, the political moderation effects we found, in addition to those found by other scholars44, indicate that psychological distance interventions may work differently across ideological groups, suggesting the need for more nuanced theoretical models.

Furthermore, there is a need to examine how these interventions impact climate-related civic engagement and collective behavior, the duration of these emotional effects, and how they relate to the three other dimensions of psychological distance, namely, social, temporal, and hypothetical5,42,61. A key limitation of our study is the reliance on student samples, which may not fully represent the diversity of the general population. Future research should address this limitation by expanding the sample to include a more diverse range of participants from different age groups, socioeconomic backgrounds, as well as personal and cultural contexts as research has illustrated that personal characteristics such as empathy and holistic and analytical thinking influence the efficacy of psychological distance interventions on climate change outcomes61,62.

These results carry significant implications for climate communication efforts by policymakers, journalists, and organizations aiming to highlight climate change as both a societal and individual threat. VR has the potential to foster adaptive emotional responses to climate change, enhance climate risk perceptions, and encourage greater investment in climate storytelling.

Methods and materials

Experimental procedure

The experiment followed a between-subjects design and comprises three main components: (i) a VR or desktop still images control condition where participants observed either a proximal or distant location while listening to a news story of climate change-related events in that location, (ii) a storytelling task in which participants described the impacts of climate change on a person from the location they viewed, and (iii) a survey assessing climate emotions and risk perceptions. All protocols were carried out in accordance with current guidelines and regulations. The methods were approved by Stanford University’s Institutional Review Board and we obtained informed consent from all the participants.

Virtual reality task

Participants were randomly assigned to one of four conditions: (i) VR with a proximal location (n = 39), (ii) VR with a distant location (n = 41), (iii) static images with a proximal location (n = 39), or (iv) digital static images on a computer with a distant location (n = 42). As recommended by prior research, participants first received an introduction to the Meta Quest 2, the headset that was used to run the experiment, and Fly VR, the VR experience platform, before engaging with the simulation34. In the proximal condition, participants chose from a list of six US locations within their Census Bureau division, selecting the one closest to their home state or the one they felt most connected to. In the distant condition, participants chose a US location they were geographically furthest from their home state.

Following this, participants accessed FlyVR, which allowed them to explore their college campus from various altitudes. After this familiarization task, participants were asked to identify their home state and its associated US Census Bureau division.Footnote 1 In the proximal and VR condition, they chose from a list of six locations for each Census Bureau division, they selected the one closest to their home state or the one they felt most connected to, and then explored this location in. In the VR and distant location conditions, participants chose a location they were geographically furthest from their home state and navigated to it in Fly VR (Fig. 4). Across VR conditions, participants were asked to take 2 min to fly around the location of their choosing at 1200 feet altitude. They were invited to look at the buildings, the water, and landmarks. Then, they were asked to fly at 400 feet altitude and explore for another two minutes before heading back to 1200 altitude and stopping. They then turned on their audio devices and listened to the news report of the corresponding area while looking around the area they were in (Fig. 5).

Participants listened to a 2-minute climate news report detailing a flood event in the location they were observing. These news stories, sourced from Associated Press coverage, and later converted to audio files read to participants via text-to-speech, were edited for consistency and included (i) a description of the location, (ii) a description of a local individual who was interviewed as part of the news coverage, and (iii) an explanation of how climate change contributed to increased flooding. Flooding is the experience selected for all participants as it is an especially relevant concern for urban areas and human-induced climate change has increased the frequency and intensity of precipitation, leading to higher flood likelihoods across the globe35,36. Following listening to the news report, participants were able to explore the area at their own pace and altitude for another 3 min.

Similarly, participants in the no VR and proximal condition selected a location close to their home state and viewed an image of it, while those in the no VR and distant condition chose a location far from their home state and viewed an image of that location as they would in a news article. Images for the control condition were selected based on the Fly VR image that participants viewed when they first loaded into the location in VR (Fig. 6).

We controlled for the length of the narrative, with each being 2 min long. Transcripts of stories heard by participants are attached in Supplemental Information 1.

Fig. 4
figure 4

Virtual reality experiences allow participants to view a distant or local location while immersed in the location (pictured: VR experience of New Orleans, Louisiana).

Full size image
Fig. 5
figure 5

Sample VR flight path images from the FlyVR experience (pictured: New Orleans, Louisiana). These images illustrate the exploration component of the VR condition, showing multiple perspectives from participants’ virtual journey through a location impacted by climate change. Participants were free to navigate through the environment with six degrees of freedom while listening to a climate news story.

Full size image
Fig. 6
figure 6

Sample CGI images of various US locations in the control condition given to participants (top left: North Shore, Massachusetts; top right: Miami, Florida; bottom left: San Francisco, California; bottom right: Nashville, Tennessee). This was the starting point for all VR experiences and participants were able to navigate through these locations in 6 degrees of freedom using hand controllers in VR. There were no visuals of sea level rise data or projections in the VR simulation and all information on flooding was covered by the audio news story only.

Full size image

Storytelling task

Following the VR or control condition, participants completed a climate storytelling task to measure how invested participants were in sharing their experience, and the number of negative emotional words in their stories. The prompt, adapted from Reynante et al. (2024)37, asked participants to describe a person from the city they experienced, who is particularly at risk from climate change.

Psychological Distance. Psychological distance was calculated as a manipulation check using a natural language processing method that combined the Linguistic Category Model (LMC), which either excludes nouns or assumes them to be highly abstract, and WordNet, which has a robust taxonomy primarily for nouns, to overcome their respective limitations63. The combined algorithm starts with the LCM equation:

$$\:\text{L}\text{C}\text{M}\hspace{0.17em}+\hspace{0.17em}\text{W}\text{N}\text{A}\:=\:\frac{\left(DAVx\:1\right)+\left(iav\:x\:2\right)+\left(SV\:x\:3\right)+\left(ADJ\:x\:4\right)+\:\sum\:WNA}{DAV+IAV+SV+ADJ+NOUN}$$

Investment. Investment was calculated by counting the length of the story that participants wrote for the storytelling task38.

Negative emotional words. Negative emotional words were calculated using the proportion of Linguistic Inquiry and Word Count (LIWC) dictionary words of negative feeling and negatively valanced words by the total number of words in the narrative39.

Survey questionnaire

After completing the storytelling task, participants filled out a survey that assessed (1) climate emotions, (2) place attachment, (3) awe, (4) spatial presence, and (5) climate risk perceptions. Demographic data, including information about participants’ home city and country, the location they viewed, and their familiarity with the location were also collected.

Political Orientation. Political orientation was assessed using a 0-100 sliding scale where 0 = Extremely liberal and 100 = Extremely conservative40. Participants were also allowed to select an N/A option.

Climate Emotions. Participants rated their emotions on a 0-100 sliding scale, where 0 indicated no emotion and 100 indicated extreme emotion. The emotions measured included: (i) depressed, (ii) miserable, (iii) anxious, (iv) afraid, (v) angry, (vi) frustrated, (vii) guilty, (viii) helpless, (ix) indifferent, and (x) hopeful27,41,42 (α = 0.91–0.95).

Place Attachment. Place attachment was assessed using a single-item, 5-point Likert scale (α = 0.92–0.93)43,44,45.

Awe. Awe was measured using an 8-item, 7-point Likert scale adapted from van Elk et al. (2016) (α = 0.83).

Spatial Presence. Spatial presence was assessed using a 3-item scale developed by the Virtual Human Interaction Lab (2021)46. Participants rated the extent to which they felt: (1) as if they were inside the virtual world, (2) as if they were visiting another place, and (3) as if they could physically interact with the virtual environment (1 = Not at all, 5 = Extremely) (α = 0.91).

Climate Risk Perceptions. Risk perceptions were measured using a 9-item index that evaluated participants’ concern about climate change, perceived likelihood of related events, perceived severity of climate change, and its global implications. Responses were provided on a 4-point scale (1 = None, 5 = Very) and calculated by summing across 9-items (α = 0.94)47.

Participants

Participants were 198 university students enrolled in a 10-week course about VR. At the start of the course, students were invited to participate in a study of how VR learning influences individual and group behaviors. All students who were part of the course took part in all the activities; however, we only included the data of those who consented to participate in this study after signing a consent form authorized by the university’s IRB and an additional oversight organization for students. To eliminate any potential perceptions of coercion, the procedure ensured that researchers and teaching team members were unaware of the identities of participating students until after the course finished by using a third-party arbiter. Of the 198 student participants who participated in the course, 163 consented to participate in the study.

The 163 participants who participated in this study (M = 78, F = 81, Non-binary or third gender = 3, Decline to Respond = 1), were between 18 and 32 years of age (M = 20.74, SD = 1.85) and identified as Asian-American or Asian (n = 68), Indigenous/Native American, Alaska Native, or First Nations (n = 1), White (n = 38), Middle Eastern (n = 4), Native Hawaiian or other Pacific Island (n = 1), African, African-American or Black (n = 15), Hispanic or Latinx (n = 15), more than one race (n = 19), other not listed (n = 1) and decline to answer (n = 1). Taking the median value of self-reported political orientations, participants identified as conservative (n = 68), liberal (n = 73) and did not answer (n = 22). Participants reported varying levels of experience with VR with 27 people having never experienced any form of extended reality technology including VR, 68 having had 1–2-time use, 28 having had 3–5-time use, 12 having 6-10-time use and 25 with more than 25 times of XR use and 5 declined to answer.

A post-hoc power analysis for a linear mixed effects model with fixed and interaction effects was run on G*Power shows that this sample size with an N = 163, an effect size of 0.25, error probability of 0.05, for 4 groups reveals a power of 0.76, indicating that there is a relatively high likelihood of detecting a true effect. This effect size was selected based on meta-analyses done on studies in climate communication and virtual reality indicating small effect sizes for media comparisons71.

Hardware

Participants were provided with Meta Quest 2 headsets (standalone head-mounted display with 1832 × 1920 resolution per eye, 104.00 horizontal FOV, 98.00 FOV, 90 Hz refresh rate, and six-degree-of-freedom inside-out head and hand tracking, 503 g) and two hand controllers (126 g) for use in their personal environment.

Virtual environment: fly VR

The session was hosted in Fly VR, a VR application powered by Google’s 3D Maps Tiles. In the application, participants fly around various locations to see the places that they were directed to at different altitudes.

Data analysis

Linguistic Inquiry and Word Count (LIWC) software was used to compute the number of words which served as a proxy for investment and the percentage of negative emotional words38.

Linear mixed models were run for each of the outcome variables where VR and spatial boundary condition, along with political orientation were treated as fixed effects, with by-participant random intercepts which included their home city. The city participants saw were initially treated as a fixed effect, but upon seeing no differences in the stimulus sample, the equation was removed. For outcome variables where the data were not normally distributed, namely, risk perceptions, climate emotions, place attachment, awe, spatial presence and negative emotions in storytelling, square root transformations were applied to fit model assumptions and then reported. All models were fitted to the data in R using lme448 with restricted maximum likelihood estimation, incomplete data were treated as missing at random, and statistical significance evaluated at alpha = 0.05. Initial models incorporated the city seen by participants and group as a random effect in the model, however, there were no significant effects of the variable on the models and thus was removed.