TL;DR
This study uses 360-degree VR technology to quantify how the layout of urban rooftop gardens affects our mental state. The researchers found that "inward-facing" designs—characterized by large floor areas, abundant greenery, and a seated perspective that obscures the surrounding city—significantly boost feelings of comfort and relaxation compared to street-facing views.
What Problem Does This Paper Address?
As urban centers like Tokyo’s Shibuya district undergo massive redevelopment, rooftop gardens are becoming vital "vertical" public spaces. However, most existing research focuses on the quantity of green space (e.g., total leaf area) rather than the quality of the spatial experience.
Architects and planners lack concrete data on how specific design choices—such as the orientation of a bench, the height of a fence, or the ratio of visible sky to buildings—impact a visitor's psychological well-being. This paper seeks to bridge the gap between architectural "spatial composition" and human "psychological evaluation" to provide a baseline for better urban landscape design.
Methods
The researchers employed a multi-step experimental approach:
- Site Selection & Categorization: They surveyed rooftop gardens near major stations on the JR Yamanote line (Shibuya to Shinjuku). Using cluster analysis on variables like floor area, plant types, and distance from the station, they selected four representative facilities (A, B, C, and D).
- Immersive VR Capture: Instead of static 2D photos, they used a 360-degree camera (THETA Z1) to record video from both sitting (110cm) and standing (160cm) eye levels.
- Psychological Testing: 21 subjects (architecture students) viewed these scenes using an Oculus Quest 2 VR headset. They rated each scene using the Semantic Differential (SD) method, evaluating 14 adjective pairs (e.g., "Closed–Open," "Artificial–Natural," "Uncomfortable–Comfortable").
- Statistical Analysis: The team used Factor Analysis to identify the underlying dimensions of the participants' feelings and Cluster Analysis to group the 12 different VR viewpoints into four distinct spatial "types."
Datasets
The study utilized a primary dataset generated by the authors:
- VR Imagery: 12 high-resolution 360-degree videos captured across four commercial rooftop gardens in Tokyo.
- Survey Data: Psychological ratings from 21 participants across 14 qualitative dimensions.
- Spatial Metrics: Pixel-level analysis of the "view area ratio" for sky, buildings, greenery, and floor for each VR scene.
Key Insights and Ideas
- The Three Pillars of Rooftop Psychology: The factor analysis revealed that three main "axes" define our experience of a rooftop: Comfort (is it cozy?), Openness (is it cramped?), and Naturalness (does it feel like a forest or a concrete slab?).
- The "Sitting Advantage": One of the most interesting findings was that sitting down (a lower eye level) generally led to higher psychological scores. Why? Because from a seated position, the garden’s own plants and fences are more likely to block out the "visual noise" of the surrounding city (roads, trains, and distant skyscrapers), creating a more immersive, private-feeling sanctuary.
- Inward vs. Outward Facing: "Inward-facing" gardens (where the viewer looks toward the center of the lush rooftop) were rated as significantly more comfortable than "street-facing" views. Even if a rooftop has many plants, if the viewer is forced to look at a busy intersection or a wall of nearby buildings, the "comfort factor" plummets.
- The "Artificial Nature" Paradox: Even though these are highly engineered spaces, the most "natural" feeling scenes were those where the spatial composition felt "enclosed" by greenery, mimicking a forest clearing rather than a high-altitude observation deck.
Key Results
- Top Performer (Group 2): VR scenes from Facility B (seated) were the most liked. These featured a high ratio of greenery, visible sky, and a "cradled" feeling created by the garden's layout. Participants described these as "forest-like" and "calming."
- Bottom Performer (Group 3): Scenes from Facility C (standing) were the least liked. These had high building-to-greenery ratios and visible urban infrastructure like fences and roads. Participants rated these as "too urban," "noisy," and "unsettling."
- Quantitative Thresholds: The study found that when the floor area is large and plants are the dominant foreground element, the "Comfortability" factor is significantly higher. Conversely, when the floor area in the view is small and dominated by artificial structures, comfort scores drop below the mean.
Limitations and Future Work
The authors acknowledge several limitations that provide a roadmap for future research:
- Homogeneous Sample: The study only tested 21 male architecture students. Future work needs to include a more diverse demographic (women, children, the elderly) to see if psychological needs differ.
- Visual Only: The VR experiment was silent. In reality, the sound of wind in the leaves or the hum of city traffic plays a massive role in how we perceive a space.
- Time of Day: The study only looked at daytime views. Given Tokyo’s famous nightscapes, the psychological impact of rooftop gardens at night is a major unexplored area.
- Environmental Factors: Variables like wind speed and temperature were not controlled but are critical to actual "comfort" in an outdoor rooftop setting.
Summary updated