Research on the Comprehensive Noise Reduction Effectiveness of Plant Communities in Urban Green Spaces

Abstract

Noise pollution significantly impacts people's health and quality of life, while urban green spaces not only beautify the environment but also play a crucial role in noise reduction and sound absorption. This study, using the plant communities of waterfront green spaces along Suzhou Creek in Shanghai as a case, innovatively develops a comprehensive evaluation system for physical and psychological noise reduction effectiveness, with 13 indicators. Through field measurements and surveys of 14 urban green spaces with different plant community characteristics, the study found that physical noise reduction effectiveness is closely related to plant species richness and vertical canopy density. Densely planted trees with thick trunks and tall shrubs significantly improve physical noise reduction effectiveness. When the vertical canopy density is less than 35%, the physical noise reduction effectiveness of waterfront plant communities (with 70% vegetation cover and 30% water cover) outperforms that of terrestrial plant communities (with no water cover) at the same canopy density level. For psychological noise reduction, increasing shrub crown width, height, and spacing improves visual perception, while enhancing vertical canopy density enriches auditory perception—both of which raise the auditory annoyance threshold. The most effective plant community combination for comprehensive noise reduction is a mix of densely planted trees, shrubs, and ground cover. The noise reduction effectiveness of a waterfront plant community is impacted by vertical canopy density. When it reaches 40% or more, the psychological noise reduction effectiveness will improve as the density increases.This study provides theoretical support and practical guidance for addressing noise issues in urban green spaces and optimizing plant community configurations.

Graphical abstract

Keywords

Urban Waterfront Green Space / Physical Noise Reduction / Psychological Noise Reduction / Comprehensive Noise Reduction Effectiveness / Plant Community Configuration

Highlight

	Competing interests  The authors declare that they have no competing interests.
	· Evaluates plant communities' comprehensive noise reduction capacity and quantifies psychological effect
	· Explores differences between terrestrial and waterfront plant communities, showing ways to reduce noise
	· Provides recommendations for plant community configuration of urban green spaces to mitigate traffic noise

1 Introduction

Noise pollution is a globally recognized environmental issue. The Environmental Noise Guidelines for the European Region issued by the World Health Organization in 2018 identifies environmental noise as the second-largest environmental stressor after air pollution, significantly affecting human health, well-being, and quality of life^[1]. Traffic noise, accounting for over 60% of urban noise, is a major contributor^[2]. Despite the inclusion of noise in China's national regular environmental monitoring since 1980, mitigation efforts in some cities have fallen short^[3]. In 2024, noise pollution remains a primary ecological environmental complaint^[4], posing challenges due to its wide impact, long-time resolution, and complexity.

Urban green spaces serve as key recreational areas while also regulating microclimates, mitigating urban heat island effect, and reducing noise^[5]. Compared with conventional engineering solutions such as noise barriers, vegetation-based noise mitigation offers a low-cost, carbon-reducing alternative^[6]. Through sound reflection and absorption, plant communities provide physical noise reduction while also generating masking effects and enhancing psychological comfort^[7]. However, most related research focuses on either physical or psychological noise reduction alone^[8][9].

This study evaluates the comprehensive noise reduction effectiveness of 14 urban green spaces with varying plant communities along central Suzhou Creek, Shanghai by developing an evaluation framework which integrates physical and psychological dimensions. The framework quantifies the comprehensive noise reduction effectiveness and examines how plant community characteristics influence different aspects of noise mitigation. The findings can support optimal plant community configurations, providing insights and practical guidance for improving plant landscapes in urban green spaces.

2 Theoretical Background

Plant communities, often referred to as "green silencers, " can absorb and block noise. The trichomes and stomata on leaves contribute to noise absorption^[10]. As traveling through greenbelts, sound waves undergo significant attenuation. With mid-to-high-frequency sounds being absorbed and scattered by tree trunks, leaves, and branches, the noise would be reduced. Thereby a well-planned combination of tree species within plant communities can effectively mitigate noise^[11]~[13].

Current research on the physical noise reduction effectiveness of plant communities primarily falls within the fields of acoustics and plant physiology, focusing on the interactions between sound waves and different mediums during propagation^[14]. Studies indicate that for single-species plant communities, key factors impacting noise reduction include plant type, physical characteristics, and vegetation quantity^[15]. Among different plant types, trees demonstrate the greatest effectiveness in noise reduction^[16], with shrubs showing comparable performance^[17], whereas lawns are significantly less effective in reduction^[18][19]. Additionally, plant density, height, and width influence noise attenuation^[20][21]. Broadleaf trees generally outperform coniferous trees in unit noise reduction effectiveness^[22] due to their denser foliage, which increases overall noise absorption capacity^[23]. Regarding compound plant communities, planting density, arrangement, and trunk diameter significantly affect noise reduction in fixed-width greenbelts^[13][24][25], with tree–shrub–herb combinations proving to be the most effective^[16].

Beyond physical noise reduction, plant communities also contribute to psychological noise mitigation through acoustic masking effects. Acoustic masking occurs when the presence of a specific sound or a visual focus reduces the perception of other sounds^[7][26]. For example, plant communities can physically shield noise sources (e.g., roads, vehicles), reducing direct exposure and increasing tolerance to noise by diverting attention^[27]. Moreover, aesthetically pleasing landscapes shift visitors' focus toward vegetation rather than noise^[28][29]. Natural sounds such as birdsong and rustling leaves, as white noise, can mask unwanted sounds and alter the perception of loudness and information content^[30][31]. Research on the psychological noise reduction effectiveness of plant communities remains in its early stages. Some studies have applied the membership principle in fuzzy set theory to assess how greenbelts affect psychological distress caused by traffic noise, using survey data to quantify the probability of annoyance caused by environmental noise and establish annoyance thresholds for specific areas^[32]. Others have employed socio-acoustic surveys on subjective annoyance levels from residents near airports to determine noise annoyance thresholds via fuzzy logic functions^[33]. Virtual reality combined with survey methods has demonstrated that richer vegetation colors and higher biodiversity contribute more to reducing noise annoyance^[34]. GIS-based studies have shown that the normalized difference vegetation index (NDVI) and green space coverage influence residents' perceived green space levels, which in turn affect their noise annoyance levels^[35].

In summary, while existing studies have explored both the physical and psychological noise reduction effectiveness of plant communities, their comprehensive noise mitigation potential remains underexplored due to the complexity and diversity of plant communities.

3 Research Methodology

3.1 Indicator Selection

Based on previous studies and relevant guidelines, this study summarized key plant community characteristics related to noise reduction effectiveness and introduced composite factors. It innovatively established a measurement and evaluation framework, at both physical and psychological aspects, to assess comprehensive noise reduction performance.

3.1.1 Key Characteristics of Plant Communities

According to the Technical Regulations for County-Level Vegetation Diversity Investigation and Assessment^[36], vegetation surveys should document community type, stratification, height, canopy closure or coverage, disturbance level, and environmental factors. For trees, species, diameter at breast height (DBH), and height should be recorded; for shrubs, species, height, density, and crown width (coverage) should be documented. Studies have shown that tree species^[37], height^[38], DBH^[39], crown width^[40], clear bole height^[41], vertical canopy density^[42], underlying surface type^[43], and plant community composition^[44] significantly impact noise reduction. Based on differences in vegetation and water coverage (denoted as UP and UW, respectively), this study categorized plant communities into terrestrial (with no water cover) and waterfront (with 70% vegetation cover and 30% water cover) types. The selected characteristics of plant communities include (Tab.1): for trees, height (TH), clear bole height (TC), DBH (TB), and planting density (TD); for shrubs, spacing (SS), crown width (SC), and height (SH); and plant species richness (PS) and vertical canopy density (VC) as overall factors. Additionally, based on principles of noise propagation^[45] and plant-based physical noise reduction^[46], this study innovatively introduced two compound factors: TD × TB / TC for trees (Factor A) and SH × SC / SS for shrubs (Factor B).

3.1.2 Evaluation Framework for Comprehensive Noise Reduction Effectiveness

(1) The evaluation indicator for physical noise reduction

The study employed the equivalent continuous A-weighted sound pressure level (LAeq) to characterize the acoustic environment. This metric, standardized according to the international standard for soundscape^[47] and China's Environmental Quality Standard for Noise (GB 3096–2008)^[48], quantifies noise exposure and is widely applied in acoustics research. Studies suggest that plant communities reduce noise through both distance attenuation and vegetation absorption^[49]. This study controlled for factors such as measurement scale, duration, and climatic conditions, ensuring that distance attenuation does not impact noise reduction results^[17][50]. Therefore, the relative noise attenuation^[40][51]—calculated as the difference between the source sound pressure level (SPL) and the SPL after transmission through plant communities—was adopted as the indicator of physical noise reduction effectiveness.

(2) Evaluation indicators for psychological noise reduction

According to acoustic masking theory, individuals' perception, experience, and interpretation of an acoustic environment depend not only on the sound itself but also on visual factors and their interaction^[52]. Thus, this study constructed a system of psychological noise reduction evaluation indicators encompassing auditory and visual perception.

The semantic differential method is widely used to assess auditory perception^[53]. Common subjective auditory evaluation indicators include "pleasant, " "lively, " "eventful, " "chaotic, " "annoying, " "monotonous, " and "calm"^[53]. Additionally, soundscape assessments frequently consider perceived sound source type^[47], noise annoyance^[54], and quietness^[55]. After removing indicators closely tied to personal experiences and merging redundant items, this study selected 7 auditory perception indicators: quietness, harmony, liveliness, richness, pleasantness, perceived sound source type, and noise annoyance (Tab.2).

Based on visual perception theories of landscapes such as prospect–refuge^[56] and stress recovery^[57], sense of color^[58], sense of space^[42], sense of layering^[56][59], sense of atmosphere^[52], attractiveness^[58], diversity^[52], and tranquility^[59] are the common visual perception indicators influencing noise tolerance^[60]. These indicators are widely applied in landscape assessment and design. Given the specific focus on plant communities, this study refined the selection to 5 visual perception indicators: sense of color, sense of space, sense of layering, sense of atmosphere, and attractiveness (Tab.2).

3.2 Selection of the Study Area

This study focused on the waterfront green spaces along the central urban section of the Suzhou Creek in Shanghai (Fig.1). Located in the city's core, this section stretches approximately 21 km from the confluence of Suzhou Creek and the Huangpu River in the east to the Outer Ring Expressway in the west. Flowing west to east through the main urban area, this segment accommodates high-quality commercial, residential, and cultural-recreational spaces^[61] and is also a focal area for urban renewal.

The waterfront connectivity improvement project for recreation and transportation purposes in this section has been largely completed, with enhancements including increased urban road network density, expanded metro capacity, and the introduction of water-based leisure routes^[62], while these initiatives have also brought noise problems. According to Shanghai's noise mapping data, the SPLs along Suzhou Creek, due to the busy traffic, generally exceed 60 dB^[63], surpassing the daytime limit (55 dB) set for Class Ⅰ acoustic functional zones in China's Environmental Quality Standard for Noise. As a result, a diverse range of waterfront green spaces with varied plant communities has emerged and a pervasive traffic noise issue exists along both banks, providing ideal sample sites for this study.

Following a preliminary experiment, 14 sample sites were selected in waterfront green spaces along the central urban section of Suzhou Creek, representing different levels of traffic noise exposure and plant community compositions. Among them, two sites are located in Jiuzi Park, two in Butterfly Bay Park, four in Mengqing Garden, and six in Changfeng Park (Fig.1). The selection criteria for these sites are as follows.

1) Sample sites of terrestrial plant communities were adjacent to urban roads (two-way, single-lane) or internal park roads and kept at least 50 m away from intersections; and waterfront plant community sample sites were situated on the river embankments opposite urban roads. Through on-site verification, site selection was confirmed that the vehicle speeds on the roads adjacent to both types of sites did not exceed 40 km/h—according to previous studies, noise frequency variations remain relatively stable at this speed^[64]. The measured traffic noise SPL was around 60 dB. The traffic flow was steady, without heavy vehicles, and light vehicles accounted for more than 95% of the traffic. The selected roads had no honking and minimal extreme noise disturbances, minimizing variations in distance-dependent noise attenuation, SPLs, and frequency-dependent noise reduction effectiveness.

2) The selected sites covered a wide range of plant community compositions, each with a distribution area exceeding the designated sampling plot size. To minimize terrain-related variability, all sites were located on flat ground, ensuring intact community compositions and healthy plant growth.

Among the designated sites, site Q0 served as the control group (single-species herbaceous type), sites Q1 ~ Q9 represented terrestrial plant communities (tree–shrub, tree–herb, shrub–herb, and tree–shrub–herb types), and sites Q10 ~ Q13 represented waterfront plant communities (tree-herb and shrub types).

3.3 Research Methods

The experiment was conducted on December 11, 2022, from 13:00 to 17:00, a clear day with no obstructions, rain, snow, or thunderstorms, and with wind speeds below 5 m/s.

First, plant community sample sites were selected and measured. A 10 m × 5 m sampling plot was delineated at the center of each site. For terrestrial plant communities, the plot covered only the vegetated area; while for waterfront plant communities, the plot also included a 10 m × 1.5 m water surface area. Second, the plant species numbers within each plot were recorded, and a laser rangefinder was used to measure TH, TC, TB, and TD, as well as SS, SC, and SH. Third, the indicators of VC, TD × TB / TU, and SH × SC / SS were calculated.

As required by the Environmental Quality Standard for Noise^[48], sound environment measurements were conducted at all 14 sample sites. A Class Ⅰ multi-functional sound level meter (AWA6228+) and a multi-channel signal analyzer (AWVA6290L+) were used to measure the SPLs of plant communities at different locations. The instruments were calibrated before measurement, and the time weighting was set to "fast" response. Based on the definition of noise attenuation, two monitoring points (O and A) were set along the central axis of each sampling plot, both located at the midpoint of the plot's long axis and spaced 5 m apart (Fig.2). The primary noise source was traffic noise from adjacent roads. During noise measurement, the sensor was positioned at a vertical height of 1.2 m above ground level. Each monitoring point was measured 5 times, with each measurement lasting 30 s and conducted at 5-min intervals. The average of these measurements was used as the final noise level for each point (α_{O_mean} and α_{A_mean})^[65].

Using the soundwalk method^[66] combined with a questionnaire survey, the psychological noise reduction effectiveness of plant communities was evaluated, with the importance of all indicators in this study assessed to calculate their weights. The questionnaire consisted of three sections. The first section covered participant demographics. The second section focused on the perception of the study site, where participants rated their visual and auditory experiences on the sample site and identified perceived sound sources. The sound source identification section was an open-ended question, while other perception ratings were assessed by a 5-point Likert scale (1 = very dissatisfied, 5 = very satisfied). The third section addressed the importance of different evaluation dimensions, including physical and psychological noise reduction. These ratings were used to determine the weight of each indicator in the comprehensive noise reduction effectiveness of plant communities^[67].

To minimize perception bias caused by individual differences, all survey participants recruited (60 in total) were local university students and received professional training before the experiment, all with normal vision and hearing and no color blindness or hearing impairments. The noise reduction assessment was conducted simultaneously with plant community characterization and physical noise reduction measurements. Participants, guided by the researchers, performed a 5-min soundwalk around Points O and A at each site before completing the questionnaire^[68]. The average scores were used as the psychological perception evaluation results for each site.

3.4 Data Processing

3.4.1 Calculation Method for Physical Noise Reduction Effectiveness

The physical noise reduction effectiveness (F_phy) of the plant community within the sampling site is determined by the noise attenuation through the community^[69], calculated using the following formula:

(1)F_{\mathrm{p}\mathrm{h}\mathrm{y}}={\alpha }_{\text{O\_mean}}-{\alpha }_{{\mathrm{A}}_{-}\text{mean}}.

3.4.2 Calculation Method for Psychological Noise Reduction Effectiveness

Based on psychophysical methods and principles of fuzzy mathematics^[32], the concept of annoyance threshold was introduced^①. The annoyance threshold refers to the critical exposure level at which a subject begins to feel annoyed in a specific environment. The difference between this threshold and the actual noise SPL at a given point determines the psychological noise reduction effectiveness of the plant community. Levels of annoyance are divided into equal intervals, and the corresponding subjective annoyance perception membership function (F) is defined as follows:

(2)F=\frac{1}{I}+\frac{0.75}{II}+\frac{0.5}{III}+\frac{0.25}{IV}+\frac{0}{V},

① For more information, please refer to the Acoustics—Methods for the evaluation and prediction of noise annoyance [GB/T 42473-2023].

where I, II, III, IV, and V correspond to the evaluation levels "very strong noise perception, " "strong noise perception, " "moderate noise perception, " "low noise perception, " and "no noise perception, " respectively, with membership degrees of 1, 0.75, 0.5, 0.25, and 0. The membership degree of evaluation level j is denoted as μ_j. The probability of annoyance (P_i) at noise SPL i is calculated as:

(3)P_i=\frac{{\sum }_j{\mu }_j{\alpha }_{ij}}{{\sum }_j{\alpha }_{ij}},

where α_ij is the frequency of occurrence of evaluation level j at noise SPL i. Based on this, the annoyance value (EL) within the sampling site is calculated as follows:

(4)EL=\frac{{\sum }_i{L}_{ki}{P}_i}{{\sum }_i{P}_i},

where L_ki represents the noise SPL at the monitoring point k and noise SPL i (in this study, referring to Points O and A).

Finally, the psychological noise reduction effectiveness (F_psy) is calculated as:

(5)F_{\text{psy}}=EL-{\alpha }_{\text{A\_mean}}.

3.4.3 Calculation Method for Comprehensive Noise Reduction Effectiveness

Based on the importance assessment results from questionnaire participants, weight conversion was performed. To minimize the impact of individual perception bias, the mean value method^[67] was used to determine the importance of indicators. The results indicated that the importance of physical noise reduction is 54%, while psychological noise reduction is 46%. Thus, the comprehensive noise reduction effectiveness (F_total) of plant communities was calculated as follows:

(6)F_{\text{total}}=1.08F_{\text{phy}}+0.92F_{\text{psy}}.

4 Results and Discussion

4.1 Characteristics of Plant Communities in the Sample Sites

The plant community characteristics of the sample sites are shown in Tab.3 and Fig.3. In terrestrial plant communities, Q0 (herbaceous) is covered solely by Cynodon dactylon. Q1 (tree–shrub) features diverse shrubs and relatively large planting spacing. Q2 (tree–shrub) has more formal and densely planted shrubs. Q3 (tree–herb) has small and sparsely distributed trees. Both being shrub–herb types, Q4 is a neatly trimmed shrub hedge with denser spacing, smaller individual crown width, and greater height than Q5. Q6 ~ Q9 (tree–shrub–herb) share similar shrub layer characteristics but differ in tree layers—Q6 and Q8 contain only large trees, Q7 and Q9 are dominated by small trees, and Q6 and Q7 have lower planting densities compared with Q8 and Q9. For waterfront plant communities, Q10 ~ Q12 are tree–herb types, with Q10 having the sparsest and Q12 the densest tree planting. Q13 is a shrub type with large shrubs.

4.2 Analysis of Physical Noise Reduction Effectiveness of Plant Communities and Its Influencing Factors

4.2.1 Results of Physical Noise Reduction Effectiveness

The mean noise SPL at the original noise source, the mean noise SPL after noise reduction by plant communities, and the physical noise reduction effectiveness at the 14 sample sites were listed in Tab.4. Results showed that among all sites, Q9 exhibited the highest physical noise reduction effectiveness, reducing noise by 11.34 dB, followed by Q4 and Q5. Sites Q7, Q13, and Q8 had moderate effectiveness, with reductions of approximately 6.5 dB. The least effective was the control group Q0, which achieved only a 1.76 dB reduction.

4.2.2 Tree-Layer Factors on Physical Noise Reduction Effectiveness

Correlation analysis of Q1 ~ Q3 and Q6 ~ Q9 using SPSS software revealed that tree-layer physical noise reduction effectiveness was significantly positively correlated with planting density (p = 0.010), but showed no significant correlation with height (p = 0.709), clear bole height (p = 0.982), or DBH (p = 0.062). However, it was positively correlated with the Factor A (p = 0.017), indicating that greater vertical obstruction by plants along the noise path can enhance physical noise reduction effectiveness.

4.2.3 Shrub-Layer Factors on Physical Noise Reduction Effectiveness

Analysis for Q1, Q2, and Q4 ~ Q9 showed no significant correlation between shrub-layer noise reduction effectiveness and spacing (p = 0.243), crown width (p = 0.852), or height (p = 0.201). However, noise reduction was significantly positively correlated with the Factor B (p = 0.017), suggesting that shrubs with greater height and crown width, coupled with smaller spacing, exhibited stronger physical noise reduction effectiveness.

4.2.4 Overall Factors Affecting Physical Noise Reduction Effectiveness

To exclude the impact of non-plant factors (water surface proportion), 10 sites with no water cover (Q0 ~ Q9) were analyzed for overall factors of plant community affecting physical noise reduction effectiveness. Correlation analysis using SPSS showed significant relationships between physical noise reduction effectiveness and both plant species richness (p = 0.033) and vertical canopy density (p < 0.001), with the latter being more significant. This suggests that higher vertical canopy density and species richness can improve physical noise reduction effectiveness.

A comparative analysis between sites with no water cover (Q0 ~ Q9) and 30% water cover (Q10 ~ Q13) generated scatter plots of vertical canopy density versus physical noise reduction effectiveness for these two plant community types (Fig.4). A linear regression analysis revealed that the slope of terrestrial plant communities (k = 0.136) was greater than that of waterfront communities (k = 0.0313), indicating that the physical noise reduction effectiveness of waterfront plant communities rose slower than the terrestrial communities as the unit vertical canopy density grew. Further calculations showed that when the canopy density was below 35%, waterfront plant communities were more effective in reducing noise; above 35%, terrestrial plant communities performed better. This suggests that for plant communities with a vertical canopy density under 35%, incorporating a water surface as the underlying surface will enhance the physical noise reduction effectiveness, while for those exceeding 35%, increasing canopy density is more beneficial.

4.3 Psychological Noise Reduction Effectiveness of Plant Communities and Its Influencing Factors

4.3.1 Psychological Noise Reduction Effectiveness of Plant Communities

A total of 60 questionnaires were distributed, with 59 valid responses. The internal consistency of the questionnaire was tested through Cronbach's α. Using the Analytic Hierarchy Process, the weight of indicators for psychological noise reduction effectiveness was determined, and the overall scores for both visual and auditory perceptions were calculated. Based on the noise annoyance results, the psychological noise reduction effectiveness at the 14 sample sites was calculated (Tab.5).

For visual perception, waterfront plant communities provided a higher overall score (M = 3.60) than terrestrial plant communities (M = 3.48). Among individual sites, Q8 scored the highest, likely due to the vertically structured trees and lush shrubs creating strong senses of layering and atmosphere. Q10 followed, with open water and Salix babylonica enhancing similar senses. Q13, dominated by dense Nerium oleander, offered a relatively pleasing visual experience, but lacked senses of color and space compared with Q10. The lowest visual score was for Q7, which had monotonous and oppressive plant colors, weak layering and spatial senses, and lacked attractiveness.

For auditory perception, Q13 had the highest overall score, and Q0 the lowest. Specifically, Q2, Q10, and Q11 were the quietest, while Q13, Q2, and Q9 had the most harmonious sounds. Q8, Q3, and Q4 were the liveliest, whereas Q8, Q3, and Q12 provided the richest auditory perception. The most pleasant auditory environments were at Q13, Q8, and Q11, whereas the highest noise annoyance values were recorded at Q0, Q3, and Q1. Additionally, birdsong was heard at Q2, Q3, Q4, Q8, Q9, Q12, and Q13, human activity and conversation sounds at Q3, Q4, Q7, Q8, and Q12, and water sounds at Q10 ~ Q13.

Among the 14 sample sites, Q9 and Q4 exhibited the highest psychological noise reduction effectiveness, followed by Q5 and Q6. The lowest effectiveness was at Q0, with Q3, Q1, and Q11 also performing poorly. For terrestrial plant communities, Q9 was the most effective due to its rich tree–shrub–herb structure and diverse colors, creating a favorable space and visual environment for birds and human activities, thus achieving a higher visual score. Q4 was favourable for its high auditory richness and low noise annoyance. Q5 followed, with balanced scores across all indicators and similarly low noise annoyance. Among waterfront plant communities, Q13 and Q12 had the best psychological noise reduction effectiveness. Q13 featured neatly arranged tall shrubs, while Q12 was dominated by Metasequoia glyptostroboides, both benefiting from water and bird sounds that enriched the auditory perception. In contrast, Q10 and Q11 had weaker psychological noise reduction effectiveness, mainly due to their lower liveliness and richness that led to poor auditory perception scores.

4.3.2 Visual Perception Influencing Factors

Correlation analysis of the visual perception scores and characteristics of plant communities showed that, for terrestrial plant communities containing shrubs (Q1, Q2, Q4 ~ Q9, Q13), shrub crown width significantly impacted visual perception. This factor had strong positive correlations with the site's senses of color (p = 0.003), space (p = 0.044), and layering (p = 0.003), and attractiveness (p = 0.009). Shrub height had the strongest positive impact on attractiveness (p = 0.008), and also impacted the senses of space (p = 0.047), layering (p = 0.017), and atmosphere (p = 0.012) to some extent. Shrub spacing was positively correlated with the sense of space (p = 0.039) and attractiveness (p = 0.045). Moreover, the Factor B positively affected the senses of layering (p = 0.006) and color (p = 0.024), and attractiveness (p = 0.022). These findings suggest that overly short shrubs, small canopies, or overly dense planting lead to relatively poor visual perception, thereby reducing the psychological noise reduction effectiveness of the plant communities.

4.3.3 Auditory Perception Influencing Factors

An analysis of auditory perception scores and plant community characteristics across the 14 sample sites revealed a positive correlation between the vertical canopy density and auditory perception richness (p = 0.043), as well as the overall score (p = 0.042). This suggests that higher vertical canopy density can better accommodate the needs of bird habitats and human activity spaces, enhancing the perception of natural and human-generated sounds, which in turn strengthens the comprehensive auditory perception evaluation.

4.4 Influencing Factors on Comprehensive Noise Reduction Effectiveness

The comprehensive noise reduction effectiveness of plant communities at all sample sites is shown in Tab.6. Among them, Q9, Q4, and Q5 exhibited the highest effectiveness, followed by Q7, Q13, and Q8, while Q0, Q3, and Q1 were the least effective.

For terrestrial plant communities, except for Q6, the physical noise reduction effectiveness generally exceeded the psychological effectiveness, and the variation trends of physical, psychological, and comprehensive noise reduction were similar (Fig.5). Q9 had the best comprehensive effectiveness due to its high vertical canopy density and dense planting, forming an effective physical noise barrier. Additionally, its strong sense of color contributed to high psychological noise reduction effectiveness. Q4 and Q5 followed, with high canopy density ensuring effective physical noise reduction and low noise annoyance. Q7 and Q8 ranked next, both featuring high vertical canopy density; Q7 had lower noise annoyance, while Q8 benefited from better senses of space, layering, and atmosphere created by the plant enclosures and higher auditory perception richness. Q0 and Q3 exhibited the lowest comprehensive noise reduction effectiveness. Q0, featuring only ground cover with the lowest vertical canopy density, demonstrated the negative psychological noise reduction effectiveness due to the worst overall auditory perception and the highest noise annoyance. Q3 showed relatively low vertical canopy density with only tree and ground cover layers, along with smaller tree DBH, resulting in the poorest auditory harmony. Q1 and Q6 also had weaker noise reduction effectiveness—Q1 suffered from high noise annoyance and a low overall score of auditory perception, whereas Q6 had poor physical noise reduction effectiveness, which impacted its comprehensive performance.

For waterfront plant communities, the comprehensive noise reduction effectiveness showed a certain correlation with the vertical canopy density (Fig.6). As the vertical canopy density of Q10 ~ Q13 increased, the visual transparency decreased, and the comprehensive noise reduction effectiveness first declined and then rose. Specifically, the physical noise reduction effectiveness increased with rising vertical canopy density. However, in terms of the psychological effectiveness, when the vertical canopy density was 10% ~ 40%, visual transparency was favourable, increased the auditory annoyance threshold, and reduced noise annoyance. When the vertical canopy density reached 40% or more, the psychological noise reduction effectiveness improved as canopy density increased.

5 Conclusions and Recommendations

This study examined the waterfront green spaces surrounding the central urban section of the Suzhou Creek in Shanghai, establishing an innovative evaluation framework with 13 indicators to assess the comprehensive noise reduction effectiveness of plant communities from both physical and psychological perspectives. The study selected 14 urban green space sites with different plant community characteristics and used noise attenuation values from field measurements along with visual and auditory perception data from surveys as dependent variables. A model was proposed to explore the effects of single and compound plant structure factors related to tree, shrub, and ground cover layers on physical, psychological, and comprehensive noise reduction effectiveness.

The results indicate that in the direction of noise propagation, richer plant species and denser vertical canopy can significantly enhance physical noise reduction effectiveness. Densely planted trees with low clear bole height and large DBH, along with tall-big shrubs, improve physical noise reduction. When the canopy density exceeds 35%, increasing it further enhances overall physical noise reduction; however, when below 35%, increasing the ratio of water surface and vegetation coverage is more effective. For psychological noise reduction effectiveness, increasing shrub crown width, height, and spacing can effectively strengthen visual perception, raising the auditory annoyance threshold. Additionally, higher vertical canopy density enriches auditory perception and the overall experience. Among all sites, plant communities combining densely planted trees, shrubs, and ground cover exhibited the highest comprehensive noise reduction effectiveness. For waterfront plant communities, variations in vertical canopy density affected visual transparency, which influenced comprehensive noise reduction effectiveness. When the canopy density is at least 40%, richly layered plant communities contributed more significantly to psychological noise reduction; below 40%, reducing it will improve visual transparency and lower noise annoyance.

Understanding the relationship between plant communities and noise reduction from a perspective of plant configuration is crucial for improving urban green spaces. Based on the findings, the results can be applied to mitigating traffic noise in various urban settings. Considering the current conditions of green spaces, the following recommendations are proposed.

1) For terrestrial green spaces with sufficient planting area, a community planting model combining densely planted trees, tall shrubs, and ground cover is recommended. Trees should be low-branching, dense-foliaged, and large in diameter, while tall shrubs should have broad canopies—both of them to be densely planted. Maximizing vertical canopy density can provide diverse wildlife habitats and increase natural sounds. Additionally, it is advisable to intersperse other tall shrubs with large canopies to highlight the shape of individual shrub plants, enhance visual perception, and further improve noise reduction effectiveness.

2) For terrestrial green spaces with limited planting area, a high-density shrub and ground cover combination is recommended. Shrubs should be large-canopied evergreen species to ensure year-round physical noise reduction. Scattered large-canopy tall shrubs can enhance visual experience, improving psychological noise reduction effectiveness.

3) For waterfront green spaces with sufficient planting area, a high-effectiveness noise reduction model similar to terrestrial vegetation should be adopted—densely planted trees, tall shrubs, and ground cover—with a vertical canopy density exceeding 40% to enhance both physical and psychological noise reduction.

4) For waterfront green spaces with limited planting area, where increasing vertical canopy density is not feasible, adjusting the proportion of water and vegetation cover can enhance physical noise reduction. Meanwhile, reducing plant density to improve visual transparency and optimize the river and water landscape, while maintaining vertical canopy density below 40%, can enhance psychological noise reduction.

This study has certain limitations. First, only 59 valid samples were collected in the psychological noise reduction survey, making the sample size relatively small and some conclusions preliminary. Second, the plant community patterns along the Suzhou Creek do not encompass all green space types. Future studies could expand the scope and increase the number of sampling sites for more comprehensive and precise data, improving the scientific validity and generalizability of the findings. Furthermore, although this study initially identified the influencing mechanisms of waterfront plant communities on comprehensive noise reduction effectiveness, the researched sample sites for waterfront plant communities were relatively few and this field remains exploratory. Future research should further investigate underlying mechanisms and refine the theoretical framework. Overall, this study aims to provide practical theoretical support and guidance for mitigating traffic noise in urban green spaces and creating multi-sensory healthy urban environments.

References

[1] World Health Organization. (2018). Environmental noise guidelines for the European region.

[2] Liu, K. (2009) A brief discussion on the harm and prevention of traffic noise pollution. China Science and Technology Information, ( 16), 21, 26.

[3] Sun, J. (2021) Annual report on prevention and control of noise pollution in China 2021. Environmental Economy, ( 12), 8– 9.

[4] Ministry of Ecology and Environment of the People’s Republic of China. (2024). Annual report on prevention and control of noise pollution in China (2024).

[5] Li, M. (2021). The principle and influencing factors of noise reduction by plants [Doctoral dissertation]. Harbin Institute of Technology.

[6] Li, M. , Van Renterghem, T. , Meng, Q. , Botteldooren, D. , & Kang, J. (2024) Tree species selection and planting strategy for reducing urban traffic noise. Journal of Applied Acoustics, 43 ( 4), 843– 853.

[7] Hao, Y. (2014) Urban design and planning from the perspective of "masking effects" in soundscape. New Architecture, ( 5), 36– 39.

[8] Leung, T. , Xu, J. , Chau, C. K. , Tang, S. K. , & Pun-Cheng, L. (2017) The effects of neighborhood views containing multiple environmental features on road traffic noise perception at dwellings. The Journal of the Acoustical Society of America, 141 ( 4), 2399– 2407.

[9] Watts, G. , Chinn, L. , & Godfrey, N. (1999) The effects of vegetation on the perception of traffic noise. Applied Acoustics, 56 ( 1), 39– 56.

[10] Zhang, L. , & Wang, S. (2012) Discussion of dust-retention and traffic noise reduction through road greening in urban area. Forest Engineering, 28 ( 5), 92– 94.

[11] Xu, X. (1997) Discussion on ecological function of the green land in the city populated area. Journal of Nanjing Agricultural Technology College, 13 ( 4), 36– 39.

[12] Bashir, I. , Taherzadeh, S. , Shin, H. C. , & Attenborough, K. (2015) Sound propagation over soft ground without and with crops and potential for surface transport noise attenuation. The Journal of the Acoustical Society of America, 137 ( 1), 154– 164.

[13] Ow, L. F. , & Ghosh, S. (2017) Urban cities and road traffic noise: Reduction through vegetation. Applied Acoustics, ( 120), 15– 20.

[14] Shi, M. , Cao, C. , Pan, X. , Chen, W. , Li, Q. , & Cao, B. (2021) Research progress on noise attenuation effect of urban landscape forest. Shaanxi Forest Science and Technology, 49 ( 3), 103– 107.

[15] Chen, L. , & Zhang, L. (2012) Summary and problems of studies on noise reduction by green land. Journal of Anhui Agricultural Sciences, 40 ( 4), 2233– 2235.

[16] Cheng, M. , & Ke, H. (1982) Noise attenuation by urban planting. Acta Scientiae Circumstantiae, 2 ( 3), 207– 213.

[17] Chen, Z. , Wang, X. , & Ye, W. (2003) Analysis on noise reduction effect of plant hedge. Forestry and Environmental Science, 19 ( 2), 41– 43.

[18] Yu, D. , Wang, X. , & Ma, X. (1998) Preliminary observations on the noise reduction effects of urban greening in Yinchuan City. Journal of Agricultural Sciences, 19 ( 1), 75– 78.

[19] Japanese Editorial Committee on Pollution Prevention Technology and Regulations, & Lu, S. (Eds.). (1988). Pollution Prevention Technology: Noise Edition. Chemical Industry Press.

[20] Cook, D. I., & Van Haverbeke, D. F. (1977). Suburban Noise Control With Plantings and Solid Barrier Combinations. U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station.

[21] Yang, H. S. , Kang, J. , Cheal, C. , Van Renterghem, T. , & Botteldooren, D. (2013) Quantifying scattered sound energy from a single tree by means of reverberation time. The Journal of the Acoustical Society of America, 134 ( 1), 264– 274.

[22] Dobson, M., & Ryan, J. (2000). Trees and shrubs for noise control. Arboricultural Advisory and Information Service.

[23] Yao, C. , Xu, Z. , & Wang, Y. (1999) Design and application of noise reduction green belt on the Shanghai section of Shanghai-Hangzhou Expressway. Shanghai Highways, ( S1), 25– 28.

[24] Yang, F. , Bao, Z. , Zhu, Z. , & Liu, J. (2010) The investigation of noise attenuation by plants and the corresponding noise-reducing spectrum. Journal of Environmental Health, 72 ( 8), 8– 15.

[25] Van Renterghem, T. , Botteldooren, D. , & Verheyen, K. (2012) Road traffic noise shielding by vegetation belts of limited depth. Journal of Sound and Vibration, 331 ( 10), 2404– 2425.

[26] Hao, Y. , Kang, J. , & Wörtche, H. (2016) Assessment of the masking effects of birdsong on the road traffic noise environment. The Journal of the Acoustical Society of America, 140 ( 2), 978– 987.

[27] Wang, H. (2019). Research on influencing factors and optimization strategy of acoustic environment in residential areas at Chongqing [Master’s thesis]. Chongqing University.

[28] Xu, W. , Wang, H. , Su, H. , Sullivan, W. C. , Lin, G. , Pryor, M. , & Jiang, B. (2024) Impacts of sights and sounds on anxiety relief in the high-density city. Landscape and Urban Planning, ( 241), 104927.

[29] Koussa, F. , Defrance, J. , Jean, P. , & Blanc-Benon, P. (2013) Acoustic performance of gabions noise barriers: Numerical and experimental approaches. Applied Acoustics, 74 ( 1), 189– 197.

[30] Van Renterghem, T. (2019) Towards explaining the positive effect of vegetation on the perception of environmental noise. Urban Forestry & Urban Greening, ( 40), 133– 144.

[31] De Coensel, B. , Vanwetswinkel, S. , & Botteldooren, D. (2011) Effects of natural sounds on the perception of road traffic noise. The Journal of the Acoustical Society of America, 129 ( 4), EL148– EL153.

[32] Zhang, B. , Hu, F. , Huang, Y. , & Pan, C. (1994) Attenuation of traffic noise by greenbelts and its effects on subjective response. Environmental Pollution & Control, 16 ( 1), 31– 48.

[33] Di, G., Liu, X., Li, Z., & Shi, X. (2010). Investigation on the airport noise annoyance threshold by socio-acoustic surveys. Conference on Environmental Pollution and Public Health (pp. 537–541). Zhejiang University of Science and Technology.

[34] Van Renterghem, T. , Vermandere, E. , & Lauwereys, M. (2023) Road traffic noise annoyance mitigation by green window view: Optimizing green quantity and quality. Urban Forestry & Urban Greening, ( 88), 128072.

[35] Dzhambov, A. M. , Markevych, I. , Tilov, B. , Arabadzhiev, Z. , Stoyanov, D. , Gatseva, P. , & Dimitrova, D. D. (2018) Lower noise annoyance associated with GIS-derived greenspace: Pathways through perceived greenspace and residential noise. International Journal of Environmental Research and Public Health, 15 ( 7), 1533.

[36] Ministry of Ecology and Environment of the People’s Republic of China. (2017). Technical regulations for county-level vegetation diversity investigation and assessment.

[37] Watanabe, T. , & Yamada, S. (1996) Sound attenuation through absorption by vegetation. Journal of the Acoustical Society of Japan (E), 17 ( 4), 175– 182.

[38] Fang, C. F. , & Ling, D. L. (2003) Investigation of the noise reduction provided by tree belts. Landscape and Urban Planning, 63 ( 4), 187– 195.

[39] Yuan, L. , Wang, X. , Wu, Y. , Gu, X. , & Wang, L. (2009) A study on traffic noise attenuation effect by tree belts in summer and winter. Highway, ( 7), 355– 358.

[40] Zhang, Q. , Zheng, S. , Xia, L. , Wu, H. , Zhang, M. , & Li, M. (2007) Noise-reduction function and its affecting factors of urban plant communities in Shanghai. Chinese Journal of Applied Ecology, 18 ( 10), 2295– 2300.

[41] Liu, X. , Niu, Y. , Wen, X. , Yuan, B. , & Hu, X. (2016) Investigation of traffic noise reduction on urban road by different green belts. Urban Environment & Urban Ecology, 29 ( 6), 18– 21.

[42] Shi, B. , & Qiu, R. (2021) Comparative analysis of the effect of green belt on traffic noise attenuation. Transport Energy Conservation & Environmental Protection, 17 ( 2), 89– 94.

[43] Yang, H. S. , Kang, J. , & Cheal, C. (2013) Random-incidence absorption and scattering coefficients of vegetation. Acta Acustica United With Acustica, 99 ( 3), 379– 388.

[44] Liu, L. (2012) A review of noise reduction effects of different types of urban green spaces. Technology Innovation and Application, ( 23), 123.

[45] Piercy, J. E. , Embleton, T. F. , & Sutherland, L. C. (1977) Review of noise propagation in the atmosphere. The Journal of the Acoustical Society of America, 61 ( 6), 1403– 1418.

[46] Ba, C. , Liang, B. , & Li, Z. (2012) Research progress of noise reduction by urban greening plants. World Forestry Research, 25 ( 5), 40– 46.

[47] Dunbavin, P. (2018). ISO/TS 12913-2:2018—Soundscape—Part 2: Data collection and reporting requirements—What’s it all about?. Acoustics Bulletin.

[48] Ministry of Ecology and Environment of the People’s Republic of China. (2008). Environmental quality standard for noise (GB 3096–2008). China Environment Publishing Group.

[49] Zheng, Z. , & Jia, P. (2010) Design methods of green belt for expressway noise reducing. Modern Transportation Technology, 7 ( 2), 97– 100.

[50] Ai, J. , Wei, Z. , Liu, W. , Liao, C. , & Chen, Y. (2019) Screening of noise reduction plant species and configuration modes in urban road green belts. Journal of Sichuan Forestry Science and Technology, 40 ( 4), 28– 33.

[51] Nie, L. (2019) Comparison of attenuation effect of different greening configurations on traffic noise. Anhui Chemical Industry, 45 ( 4), 96– 99.

[52] Chen, S. , He, P. , Yu, B. , Wei, D. , & Chen, Y. (2024) The challenge of noise pollution in high-density urban areas: Relationship between 2D/3D urban morphology and noise perception. Building and Environment, ( 253), 111313.

[53] Axelsson , Ö. , Nilsson, M. E. , & Berglund, B.. (2010) A principal components model of soundscape perception. The Journal of the Acoustical Society of America, 128 ( 5), 2836– 2846.

[54] Jeon, Y. J. , Lee, J. P. , Hong, Y. J. , & Cabrera, D. (2011) Non-auditory factors affecting urban soundscape evaluation. The Journal of the Acoustical Society of America, 130 ( 6), 3761– 3770.

[55] Pheasant, R. , Horoshenkov, K. , Watts, G. , & Barrett, B. (2008) The acoustic and visual factors influencing the construction of tranquil space in urban and rural environments tranquil spaces-quiet places?. The Journal of the Acoustical Society of America, 123 ( 3), 1446– 1457.

[56] Appleton, J. (1996). The Experience of Landscape. Wiley Chichester.

[57] Ulrich, R. S. , Simons, R. F. , Losito, B. D. , Fiorito, E. , Miles, M. A. , & Zelson, M. (1991) Stress recovery during exposure to natural and urban environments. Journal of Environmental Psychology, 11 ( 3), 201– 230.

[58] Li, D. , Zhang, Q. , Dong, K. , Wu, X. , & Zhou, R. (2021) Evaluation of plant landscape space in the campus of Tianjin University. Journal of Chinese Urban Forestry, 19 ( 4), 115– 119.

[59] Grahn, P. , & Stigsdotter, U. K. (2010) The relation between perceived sensory dimensions of urban green space and stress restoration. Landscape and Urban Planning, 94 ( 3-4), 264– 275.

[60] Shao, Y. , Xue, Z. , Hao, Y. , & Yin, Y. (2022) Research on evaluation of audio-visual perception quality of urban parks: A case study of Chengdu Outer-ring Ecological Zone. Landscape Architecture, 29 ( 9), 26– 32.

[61] Shanghai Municipal People’s Government. (2021). Shanghai “One River, One Creek” development plan for the 14th five-year period.

[62] Shanghai Municipal Bureau of Planning and Natural Resources. (2021). Suzhou river waterfront area construction plan (2018–2035) (public edition).

[63] Xu, X. , Ge, Y. , Wang, W. , Lei, X. , Kan, H. , & Cai, J. (2022) Application of land use regression to map environmental noise in Shanghai, China. Environment International, ( 161), 107111.

[64] Luo, P. , Cai, M. , & Wang, H. (2013) Frequency characteristics of noises of various types of vehicles in different speeds. Noise and Vibration Control, 33 ( 5), 86– 89.

[65] Tang, B. (2016) Analysis of effect of different plant communities on the noise reduction in Nanping City. Journal of Sichuan Forestry Science and Technology, 37 ( 5), 81– 83.

[66] Aletta, F. , Guattari, C. , Evangelisti, L. , Asdrubali, F. , Oberman, T. , & Kang, J. (2019) Exploring the compatibility of "Method A" and "Method B" data collection protocols reported in the ISO/TS 12913-2: 2018 for urban soundscape via a soundwalk. Applied Acoustics, ( 155), 190– 203.

[67] Deng, X. , Li, J. , Zeng, H. , Chen, J. , & Zhao, J. (2012) Research on computation methods of AHP weight vector and its applications. Mathematics in Practice and Theory, 42 ( 7), 93– 100.

[68] Bahalı, S. , & Tamer-Bayazıt, N. (2017) Soundscape research on the Gezi Park–Tunel Square route. Applied Acoustics, ( 116), 260– 270.

[69] Liu, G. , Wang, F. , & Fan, Y. (2021) Research on the noise reduction effects of plants in highway greening belt. Agriculture and Technology, 41 ( 2), 100– 102.

RIGHTS & PERMISSIONS

© Higher Education Press 2025