Rapid urbanization, while improving quality of life, negatively impacts mental and physical well-being through reduced access to nature and increased stress. Urban parks offer a refuge, providing various health and social benefits, but noise pollution is a significant concern. This study addresses the gap in understanding the integrated effects of visual and auditory environments on park visitors' experiences. Stress Recovery Theory (SRT) and Attention Restoration Theory (ART) highlight the restorative potential of natural environments, suggesting they can alleviate stress, improve emotional states (ES), and enhance perceived restorativeness (PRS). Landscape preference (LP) is also a crucial factor, as positive experiences are linked to a preference for specific environments. Existing research often focuses on single sensory impacts (visual or auditory) rather than their complex interplay. This research aims to explore the combined effects of visual and auditory environments on LP, ES, and PRS in urban parks, providing valuable insights for urban park planning and design to enhance the restorative potential of these spaces.

Landscape perception, including emotional responses, is central to landscape preference studies. Research shows strong preferences for nature-related visual elements (water, vegetation) often explained by prospect-refuge theory and evolutionary psychology. Auditory perception also influences LP, sometimes even more significantly than visual perception. The combination of specific sounds (birdsong, water) can increase immersion and preference. Visual esthetics and functional landscapes are closely linked to sound preferences. Natural landscapes positively impact psychological and physiological well-being, reducing stress, improving concentration, and reducing fatigue. Auditory perception plays a vital role in restorative benefits; high-quality sound enhances restorative environments. Interactions between visual and auditory stimuli can increase restorative value, with diverse and complex landscapes and soundscapes fostering greater psychological restoration. However, existing research predominantly focuses on individual sensory effects, neglecting the intricate connections between LP, ES, and PRS in complex audiovisual environments.

This study was conducted in five diverse urban parks in Hangzhou, China, selected to represent varied audiovisual environments. Data collection involved face-to-face surveys using paper questionnaires from September 20 to November 5, 2023, targeting visitors meeting specific criteria (normal hearing/vision, no psychological issues, minimum 5-minute stay). 861 completed questionnaires were collected (93% response rate). Demographic information (gender, age, education, occupation, park visit frequency, time spent) was recorded. Visual data were gathered using panoramic cameras and analyzed using deep learning to quantify landscape features (trees, water, sky, ground vegetation, buildings). Auditory data were collected using a binaural headset microphone and analyzed for various acoustic parameters (A-weighted equivalent sound pressure level, sound exposure level, percentile sound levels) and psychoacoustic indicators (loudness, sharpness, roughness, fluctuation strength). The survey included scales to assess landscape preference (complexity, mystery, coherence, legibility, landscape quality, willingness to experience), emotional state (calm, secure, comfortable, etc.), and perceived restorativeness (fascination, being-away, compatibility, extent). Data analysis used SPSS for preliminary analysis (Spearman correlation for non-normal data) and PLS-SEM for path analysis to explore direct, indirect, and moderating effects of audiovisual environments on LP, ES, and PRS. A second-order measurement model was developed for PRS.

Preliminary analysis revealed significant positive correlations between natural sounds (birdsong, flowing water) and positive outcomes (LP, ES, PRS) and negative correlations between human-made and mechanical sounds and these outcomes. Trees, ground vegetation, and water views showed positive correlations, while buildings and sky views showed negative correlations. PLS-SEM analysis revealed that LP, ES, and environmental acoustic indicators (EAI) significantly influenced PRS. Natural landscape features (trees, ground vegetation, water) and natural sounds (birdsong, flowing water) positively influenced PRS, LP, and ES, while artificial elements and anthropogenic sounds had negative effects. Mediation analysis indicated that LP enhances PRS through ES mediation. Natural visual elements indirectly enhanced ES and PRS by improving LP. Bird songs and flowing water sounds positively influenced emotions and PRS indirectly via LP and ES. Moderation analysis revealed that bird songs positively moderated the effects of water and ground vegetation on PRS, while traffic and construction sounds negatively moderated the relationship between building views and PRS. Bird songs positively moderated the effect of water views on ES, but negatively moderated the effect of sky views. Flowing water sounds positively moderated ES in environments with trees and water views. Conversation sounds negatively moderated ES in environments with buildings and ground vegetation but positively moderated ES with tree views. Traffic sounds negatively moderated ES in environments with ground vegetation or water.

The findings support the importance of both LP and ES in influencing PRS, with LP mediating the relationship between environmental features and PRS. The positive impact of natural elements aligns with SRT and ART. The negative impact of anthropogenic sounds and built environments highlights the need to minimize these elements in urban park design. The nuanced effects of bird songs, depending on acoustic quality, suggest the need for careful soundscape design. The contrasting effects of conversation sounds in different vegetation types highlight the complexity of audio-visual interactions. The results suggest the need for careful consideration of the composition, density, and spatial arrangement of vegetation, considering prospect-refuge theory and ensuring clear sightlines.

This study provides valuable insights into the interplay between visual and auditory environments and their impact on park visitor experience. Natural elements consistently enhance preference, positive emotions, and perceived restorativeness. Careful consideration of both visual and auditory design is crucial for maximizing the restorative potential of urban parks. Future research should investigate the impact of other senses, explore specific acoustic parameter ranges for optimal bird song integration, and examine the role of individual factors (personality, place attachment, etc.) in shaping perceptual responses.

The study's geographic focus on Hangzhou may limit the generalizability of findings to other contexts. The cross-sectional nature of the study prevents causal inferences. The use of self-reported data introduces subjective biases. Future research should consider a wider geographical scope, longitudinal designs, and incorporate physiological measures to corroborate self-reported data. Further exploration of the interplay of other senses beyond vision and hearing is also recommended.