Residents’ Preferences to Multiple Sound Sources in Urban Park: Integrating Soundscape Measurements and Semantic Differences

Abstract

Urban parks provide their residents with a space for activities and interactions, and soundscapes play an important role in influencing the residents’ sense of experience of the park environment, with the result that the preferences of residents regarding multiple sound sources can be used as the basis of improving the soundscape quality and optimizing the function of urban parks. Using Shanghai Century Park as the research area, this study focused on the residents’ preference to multiple sound sources based on registered residence differences, and comprehensively used the semantic differential method, importance-performance analysis model, and multinomial logit model to conduct quantitative research on urban park soundscapes and combined the measurement results of the soundscape with the residents’ perception and preference. The results showed that the artificial sounds in Century Park were noisy during the morning and evening peak, while the human sounds were noisy when residents gathered in the park. There was a close relationship between the favorability and subjective loudness of urban park soundscapes, and the residents’ perceptions of urban park soundscapes substantially differed in terms of loudness, tone, and sense of belonging, whereas the differences in the perception of timbre, coverage, and sound source distance were relatively small. Furthermore, it is necessary to enhance the role of natural sounds as well as control the main noise-producing sound sources so that various sounds in the park will not interfere with each other; the optimization of soundscapes should also focus on the residents’ different soundscape preferences due to their individual characteristics.

1. Introduction

As an important part of urban ecosystems, urban parks have important ecological, economic, and social values as they provide multiple material and immaterial services for urban residents. With increasing urbanization and the growing needs of urban residents regarding the ecological environment, the demand for urban parks is becoming increasingly diversified and complicated. In particular, the rapid urbanization process in China has led to problems such as parks being pushed out of the urban landscape, environmental damage, and structural singularizations, which reduce the ecological welfare of urban residents. In recent years, the Chinese government has been carrying out a new type of people-centered urbanization that emphasizes green development to make more livable cities. For example, the Development Strategy of the Yangtze River Delta and the Development Strategy of the Guangdong–Hong Kong–Macao Greater Bay Area implemented in China in recent years, which all highlight the new urbanization construction and improvement in the urban human settlement environment. With this type of urbanization in mind, it is necessary to break through the traditional model of urban park construction to ensure that parks fully embody the people-centered development concept, organically integrate public and ecological services, and enhance their ornamental and practical values.

Generally, the quality of ecological services can be reflected by the perceptions of their demand; that is, people perceive the world through their senses [1,2]. The term “soundscape” was first proposed by Finnish geographer Granoe in 1929. He believed that soundscapes should be centered on the listener, making sound a phenomenon that can be used to perceive various content [3]. That is, similar to the word “landscape”, the term “soundscape” is mainly used to describe the sound environment of the person hearing the sound. However, because most Western countries had serious noise problems at that time, most research on soundscapes was devoted to dealing with noise pollution and did not pay much attention to the functions of the soundscapes themselves. In 1969, the soundscape concept was re-emphasized through studies trying to determine how people perceived the sound environment and how soundscapes affected their view of the city [4]. In 1976, Canadian composer and ecologist R.M. Schafer formally defined the concept of “soundscape” to be the sound environment perceived and understood by individuals and society [5]. Therefore, soundscapes were not just a background component of the environment, but also how the sound space was regarded at the core of interactions. Soundscapes depend on the relationship between humans and the environment in which they are located [6]. In this study, soundscapes are the sound environments that people perceive in a given scenario. Therefore, during the study of soundscapes, more attention should be paid to individual concepts rather than physical measures, and more consideration should be given to positive sounds rather than only noise. In addition, related issues related to the emotional dimensions should be more thoroughly considered. Therefore, in order to best understand how people interact with specific environments, soundscape research should focus on the combination of three factors: the sound, listener, and environment [7].

To date, research on soundscapes has mainly focused on locations such as residential areas, commercial pedestrian streets, waiting areas in railway stations, and urban forests, and it has emphasized and discussed the advantages and disadvantages of topics such as noise pollution, the value and importance of natural silences, multi-sensory activities, and music tourism [8,9,10]. Studies have largely focused on the complex relationship between the so-called “social soundscape” produced by human society and the sum of many sounds that human beings hear in the natural environment, namely, the “natural soundscape”. For example, natural sounds such as birdsong, water, and the rustling of leaves have been highlighted as positive soundscapes, whereas traffic sounds have been shown to be negative [11,12,13]. Similarly, it was once pointed out that sound and visual stimuli are interdependent, but that they also depend on the type of sound [14]. Therefore, the quality and type of soundscapes could be key factors for creating sustainable urban parks [15,16,17,18]. Furthermore, it has also been pointed out that there is a complex interaction between noise and active natural sounds in how soundscapes are perceived and preferred. For example, if birdsong is added to the soundscape, people can perceive a loud environment more actively and cover up less pleasant noises [19]. More case studies are needed to illustrate the conclusions in this regard, but the existing studies have at least given some clues. In daily life, various types of sounds such as birdsong, music, and chatting sounds widely exist in urban areas and are mixed with each other. They together constitute the sound environment of urban areas. In short, the goal of soundscape discussions should be to attain the close combination of natural and social soundscapes and pay attention to the most practical demands and preferences of individuals. Therefore, thorough discussions should be carried out related to the harmonious relationships between soundscapes, landscapes, and people.

The cosmopolitan city of Shanghai has a complex population structure including nearly 25 million local residents and seven million migrants per year. The exploitation and use of urban resources in Shanghai are insufficient and unbalanced. Existing studies have focused extensively on the impact of individual factors on noise, subjective loudness, comfort levels, and other soundscape experiences [20,21,22,23]. The above shows that urban parks are an important type of infrastructure in urban living environments, and that soundscapes play an important role in both the external image and internal functions of urban parks. Therefore, in this study, the soundscapes of urban parks were considered as the intersection and medium for different user groups to interact and communicate with each other. It should be pointed out that the “soundscape features” this study focused on include the residents’ subjective perceptions of urban park soundscapes, the influence of the residents’ individual characteristics on their soundscape perception preferences, and the urban park soundscape preferences and their driving factors for different user groups. Before conducting a series of analyses, different sound types and their characteristics are identified, different urban residents’ groups and their characteristics, and the residents’ semantic differential perception of different sound types. Based on this, the soundscape features of a typical urban park in Shanghai and the differences among different user groups according to their residence status were analyzed (“registered residence” is a household-based population management policy implemented in China that shows the legitimacy of a resident’s residence in a certain place) including their perceptions of urban park soundscapes, in order to provide them with opportunities for better interactions and communication. Among them, the division of different user groups is mainly based on the actual population structure of Shanghai. Furthermore, the residents’ soundscape perceptions may differ among groups of users and between individuals of the same “group” as it depends on multiple factors including psychological factors, previous experiences, personal preferences, and attitudes in addition to socio-demographic variables including age, gender, education level, etc. Therefore, in the process of specific analysis, the above factors are fully considered [24,25]. Thus, the corresponding category variables were set to be obtained together with other data in the process of the questionnaire survey.

Due to the differences between Eastern and Western cultures, the design concept of soundscapes in Chinese classical gardens concerns the expression of emotions and the creation of artistic conception, whereas that of the West is more modern, diversified, and pays more attention to the practicability of soundscape construction in landscape design. Thus, the typicality of relevant research in China can best be reflected through evaluating the current research hotspots of soundscapes as well as analyzing the design and optimization of soundscapes and the individual experiences of urban parks from the perspective of promoting the perceptions of different user groups [26].

To sum up, the aims of this study were divided into three parts: the first was to discuss the residents’ subjective perceptions of urban park soundscapes, the second aim was to discuss the specific situations and directions for improvement of the soundscapes, and the third aim was to discuss the influence of the residents’ individual characteristics on their soundscape perception preferences. This study focused on the diverse subjective experience residents have of soundscapes as well as the inherent operation laws of the functional realization of soundscapes. It should be pointed out that perception refers to the residents’ subjective feelings about the types, characteristics, and realization of the functions of soundscape, while preference refers to the judgment of the residents’ preference for soundscapes based on their own needs. The key issues are as follows: (1) What aspects of soundscape features are included in urban parks, and what is the degree of different characteristics? (2) What are the main types of urban park soundscape perception experiences, and what are the main strategies for coping with and promoting them? (3) What are the urban park soundscape perception preferences of different user groups, and what are the factors influencing these preferences?

2. Materials and Methods

2.1. Study Areas and Observation Points

Shanghai has experienced a rapid urban expansion in recent years, and the urban development of Shanghai has a significant impact on the residents’ experiences in urban parks. Based on the geographical location, the landscape characteristics, population capacity, and layout of Shanghai’s urban parks, this study utilized Shanghai Century Park as the research area. Shanghai Century Park was part of the first batch of Chinese Class 4A Tourist Spots established in 2002, and it covers an area of 140.3 hectares with a total green area of approximately 860,000 m². It is the largest ecological urban park with rich natural landscapes located in the central inner ring of Shanghai. Shanghai Century Park was selected as the object of soundscape investigation for two main reasons.

Specifically, Shanghai Century Park has a good overall environment and is a comprehensive tourist attraction that includes both ecology and entertainment. As the largest ecological urban park in Shanghai, Century Park retains natural landscapes, it has a variety of biological landscape tourism resources, and its soundscape elements are particularly abundant. Second, Shanghai is located in the subtropical monsoon climate region, with a mild and humid climate, sufficient sunshine, and an annual average temperature of about 16 degrees Celsius. Unlike many regions in northern and southern China, the climate of Shanghai is warm in winter and cool in summer, which is relatively stable. This climate condition provides a relatively good basis for the activities of Shanghai residents in urban parks. Considering these two factors, Shanghai Century Park was selected as the site for soundscape field surveys and subjective and objective evaluations in this study (Figure 1).

Twelve representative samples of park areas were selected to measure their soundscapes. The areas were selected to make the sampling distribution range cover as much as the whole park as possible. Furthermore, the 12 samples were classified into three sample types: scenic spots, functional areas, and natural areas. Scenic spots included well-known tourist attractions, viewing platforms, and landmark buildings such as the Century Flower Clock, Musical Fountain, Lawn Pigeon, and Amusement Park. Such areas often provide space and objects for publicity and education, scientific research, and other activities in the park. Functional areas inside Century Park include roads, squares, entrances, and exits such as Gate 1, the Grand View Terrace, Sculpture Square, and the Fenghe Bridge. Such areas ensure the normal operation of the park, and they are thus indispensable park infrastructures. Natural areas refer to areas such as gardens, woods, and lakes including Century Plum Garden, Autumn Garden, Birds Island, and Lucky Pond. These areas constitute the main body of the park’s natural landscape, and they provide a basis for the ecological services and benefits of the park.

There are many types of sounds in Century Park. According to the sound source features, these sounds can be divided into three categories: human, artificial, and natural sounds (

Table 1. Composition of the Century Park soundscapes.

Sound Types	Content
Human sounds	Entertainment activity, playing, vendors, chatting
Artificial sounds	Traffic, entertainment equipment, music, construction
Natural sounds	Wind, birdsong, leaves, water

).

Human sounds mainly come from the various recreational behaviors of residents including the sounds of chatting, playing, and engaging in entertainment activities as well as the sounds of staff members doing their jobs. These human sounds are the most dynamic soundscapes in the park. Artificial sounds in Century Park are also a common type of soundscape, mainly including the sounds of traffic, construction, and mechanical equipment. These artificial sounds come from large-scale entertainment equipment in the park and broadcasting equipment that provides information to the residents. Sightseeing buses, tandem bikes, and other vehicles in the park are the sources of traffic sounds inside the park, and the traffic sounds generated at the entrance of the park and the nearby subway station also have some impact. In addition, there are occasionally some staff members carrying out construction work that produces numerous construction sounds such as those from mowing and logging (Figure 2). Natural sounds in Century Park mainly include the sounds of birds, leaves, wind, water, and other natural phenomena. Birdsong has the widest distribution because Bird Island, a bird reserve with rich natural landscapes, is in the central area of Century Park. Additionally, because Shanghai is in the coastal area of eastern China, its subtropical monsoon climate results in natural sounds such as wind, rain, and leaves. Furthermore, the sound of water in the park is generated by various streams and musical fountains that run at a set time (Figure 3).

2.2. Experimental and Survey Design

This study combined soundscape measurements and semantic differences methods for the analysis, which is an extension of existing research. Previous studies have also conducted similar discussions by combining quantitative, qualitative, and spatial analysis methods [27,28]. In fact, mixed methods are very common in soundscape research, and these exploratory studies provide the possibility for more detailed and comprehensive soundscape research [29,30].

From the aspect of soundscape measurements, this study introduces the concept of sound level. Sound level refers to the physical quantity consistent with people’s subjective perception of sound intensity, and the unit is dB. The measuring instrument used was a Smart Sensor Digital Sound Level Meter AS804, with a measurement range of 30–130 dB, and accuracy level of −1.5 dB to 1.5 dB (Figure 4). In specific use, we set the measurement height above the ground level to 1.5 m, and the equipment is automatically calibrated and stationary during the measurement. To describe the objective sound level of Shanghai Century Park, the sound level at 12 sample areas determined in Shanghai Century Park were measured in May 2021. Specifically, since Shanghai Century Park has opening and closing hours, the measurement time from 8:00 to 18:00 was divided into 10 time periods, and every hour was a measurement period. In each period, the duration of each measurement was 10–20 min [31]. Based on this, this study measured and recorded the sound level of each sampled area in different periods. Furthermore, since the sound level is easily affected by the objective measurement environment such as sudden noise, rain, or wind, and we finally used the average value after 10 measurements in the same sampled area, the measurement was taken on days that were calm to a gentle breeze and without rain (Appendix B). Therefore, no data were collected under the condition of wind speed exceeding 5 m/s. In the measurement process, this study refers to the relevant standards of the International Organization for Standardization (ISO) for measurement and calibration [32,33].

Subjective evaluations of soundscapes by park users can be used as an important reference for urban park soundscape construction. Previous studies have pointed out that soundscape perceptions are related to soundscape types, individual preferences, and other demographic factors [34,35]. This study conducted a questionnaire survey including the respondents’ basic information (e.g., age and gender) and other factors that have been shown to affect soundscape perceptions (

Table 2. The questionnaire variables and rating scores.

Variables	Attribute Value
(Y) Preferred soundscape types	1 = Human sounds; 2 = Artificial sounds; 3 = Natural sounds
(X1) Gender	1 = Men; 2 = Women
(X2) Age	1 = Age 17 and under; 2 = 18–29 years old; 3 = 30–39 years old; 4 = 40–49 years old; 5 = 50–59 years old; 6 = Age 60 and over
(X3) Residential area	1 = Within the Inner Loop; 2 = Inner Loop–Middle Loop; 3 = Middle Loop–Outer Loop; 4 = Outer Loop–City boundary; 5 = Outside the city border
(X4) Marital status	1 = Unmarried; 2 = Married
(X5) Family size	1 = 1 person; 2 = 2 people; 3 = 3 people; 4 = 4 people; 5 = 5 or more people
(X6) Health status	1 = Very good; 2 = Relatively good; 3 = Neutral; 4 = Relatively poor; 5 = Very poor
(X7) Residence status	1 = Native-born; 2 = First-generation immigrant; 3 = Second-generation immigrant; 4 = Foreign registered residence
Favorability	1 = Very annoying; 2 = Relatively annoying; 3 = Neutral; 4 = Relatively preferred; 5 = Very preferred
Importance	1 = Not at all important; 2 = Somewhat Unimportant; 3 = Neutral; 4 = Somewhat important; 5 = Very important
Subjective loudness	1 = Very quiet; 2 = Relatively quiet; 3 = Neutral; 4 = Relatively loud; 5 = Very loud

) [36,37,38,39,40,41]. Furthermore, this study also investigated the respondents’ perception and evaluation of the favorable, importance, and subjective loudness of the soundscape in the park. A favorable response indicated the preference for a certain soundscape type, which was used to reflect the actual performance of a specific soundscape type. The importance degree refers to the evaluation of the importance of different soundscape types, which is the result of the residents’ subjective judgments. In other words, the importance degree mainly emphasizes the residents’ subjective perception of different soundscape types rather than the objective evaluation based on the actual situation. The degree of subjective loudness refers to evaluations of loudness according to the different sensitivities of human ears to sound intensity. In this study, the Likert scale method was adopted, and the scores considered were 1, 2, 3, 4, and 5. The respondents were required to score the favorability, importance, and subjective loudness of each soundscape type. Each score represents a respondent’s feelings. Compared with other methods, the Likert scale is simpler and easier to design, and it is less difficult for respondents to fill in [42].

It is necessary to point out that different types of residents will have different soundscape perceptions and experiences in urban parks [43,44,45,46]. Since China’s Reform and Opening Up, the benefits of Shanghai in terms of economic, cultural, and educational factors as well as medical treatment have attracted and accommodated a variety of population sources. This is especially reflected in the “registered residence” of different groups of residents. “Registered residence” is a household-based population management policy implemented in China that shows the legitimacy of a resident’s residence in a certain place. The data show that in 2022, more than 40% of the population in Shanghai is composed of migrants, which in most cases do not have registered residences in Shanghai. Based on the registered residence policy and the actual composition of Shanghai residents, this study divided the respondents of our questionnaires into four types. “Native-born” indicates that the respondents and their parents were all born in Shanghai and held registered residences. “First-generation immigrant” indicates that the respondents previously registered their household registration elsewhere and later emigrated to Shanghai where they obtained registered residences. “Second-generation immigrant” indicates that the respondents’ parents were first registered elsewhere but later emigrated to Shanghai, and the respondent was born in Shanghai and automatically given household registration. “Foreign-registered residence” indicates that the respondents do not have Shanghai household registration and are immigrants to Shanghai. In this study, the number of native-born residents and foreign registered residence residents was 236, accounting for 31.51% of the total number; the number of first-generation immigrants was 100, accounting for 13.35% of the total number; and the number of second-generation immigrants was 177, accounting for 23.63% of the total number. The proportion of the above population is basically consistent with the actual population composition of Shanghai.

It should be pointed out that our survey contained two parts: one including the basic information of residents, the questionnaire of the semantic scale, and the Importance-Performance Analysis (IPA) Model questions, and the other was an experiment on the residents’ preferences for different voice types (Appendix A). All surveys were conducted in Century Park. Specifically, the three types of urban park soundscapes had different features that affected the interviewees’ response processes. Considering this situation, audio playback equipment and noise-cancelling headphones were used to show different sound types to the interviewees to obtain their preference data. This study presented the sounds in the following manner: in each audio clip, three sounds were played to the respondents, and the duration of each type of sound was about 10 s. Two sounds in the previous audio clip were replaced with two new sound types in the next clip; thus, different combinations of various sound types were created. The three kinds of sounds contained in each audio clip were isolated sound sources recorded by the interviewers in a real-life environment. During the recording process, the original characteristics and effects of each specific sound were guaranteed as much as possible, that is, they were not disturbed and affected by other sounds. During this process, interviewees were asked to rate each sound in each audio clip according to the Likert scale, with a score of 1 to 5 (1 = most least favorite; 5 = most favorite); each respondent had 30 s to make their decisions. After the survey, the interviewees received small gifts such as umbrellas and small fans. The audio clips used in this study and their scoring processes are described in Figure 5. The interviewers were responsible for collecting the data, so they knew which of the different sound types belonged to human, artificial, and natural sounds. However, the interviewees did not know this information; thus, their scores were based entirely on their own subjective preferences. The point of this procedure is to fully understand the residents’ preferences for various sounds without being affected by some familiar sounds or distinctive sounds. Meanwhile, all possible combinations for each participant were tested. This evaluation method based on the comparison of three sound types can better reflect the differences between different sound types than the method of evaluating a single sound.

In addition, considering that the subjective feelings of urban park users will be affected by factors such as age, gender, and living areas, this study attempted to have as wide a scope of interviewee groups as possible, and then we performed selective statistical analyses to ensure the effectiveness of the survey results. To ensure that the survey process was scientific, and the results were accurate, the interviewers participating in the survey were all undergraduate or graduate students with professional background knowledge and field survey experience. To improve the survey efficiency, namely, the inter-experimenter consistency, a one-day training session for all interviewers before the survey was conducted including the research background, purpose, implementation methods, inquiry methods, and other precautions. In addition, as a supplementary means of the questionnaire survey, this study also conducted interviews. For example, when interviewers need to further verify and expand the answers of the interviewees, they will use methods such as conversation and Q&A to obtain other effective information other than through the questionnaire. The reason for setting up the interview process in this study was mainly to obtain some effective information beyond the results of the questionnaire survey to combine the conclusions of the survey with the objective reality for analysis. In addition, the larger the sample size, the more representative the sample is of the population. Considering that Shanghai Century Park is in a central urban area with a large population density in Shanghai, this study selected as many samples as possible. The formal survey was conducted on 1–4 May 2021, and it took about 30 min to complete the questionnaire including nine questions such as the semantic scale and IPA questions. In addition, as the time period of the survey happens to be the Labor Day holiday in China, it can ensure that the sample size of the park is sufficient. At the same time, due to the impact of COVID-19, there is little inter provincial migration of people during the Labor Day holiday, which can ensure that Shanghai residents are the main respondents of the park. This objective situation was also confirmed in the statistical results of the sample description in this study.

Finally, a total of 800 questionnaires were completed. After excluding invalid questionnaires (mainly include incomplete questionnaires, inconsistent questionnaires, and questionnaires with obvious errors), 749 questionnaires were obtained, with an effectiveness rate of 93.62%. Therefore, the total sample size of this study was 749.

2.3. Statistical Research Methods

2.3.1. Semantic Differential Method

First, this study used the semantic differential method to analyze the characteristics of urban park soundscapes. The preference survey of the interviewees has always been an important method in urban spatial perception research [47]. The semantic differential method (SD) is a psychological research method proposed by Osgood (1957) that can be used to objectively carry out the quantitative analysis of research objects [48]. In fact, numerous studies in many research fields have adopted the SD method, and most of them have combined it with other statistical methods [49,50,51,52,53,54]. The SD method holds that human beings have extensive and common emotional meanings for concepts or words, so it is reasonable and effective to directly ask interviewees about the “meaning of a concept.” In this study, the evaluation results of urban park soundscape features were based on the perceptions of the subjects; that is, the subjective evaluations of urban park soundscape features were provided by urban residents based on their auditory senses. Some studies have pointed out that if a public space is open and accessible, the human interactions within it will help individuals participate and create a sense of acceptance [55,56]. In this study, the characteristics of urban parks such as their coverage and the distances of sound sources could have an impact on the residents’ perceptions, which in turn impact their behaviors. Furthermore, external stimuli could also trigger this interaction and enhance the interactions and familiarity between individuals [57]. This can be reflected in aspects such as the timbre, tone, and loudness of urban park soundscapes. Additionally, people’s interactions in urban public spaces are often attributed to clear reasons such as comfort levels and their sense of belonging and security [58]; thus, these factors are also important parts of the residents’ soundscape experiences. Based on this understanding, 11 pairs of adjectives with opposite meanings were selected to describe the soundscape characteristics of urban parks and constructed a semantic difference scale for the feature analysis of urban park soundscapes. On the scale, the 11 evaluation variables corresponded to the pairs of adjectives, which were the evaluation indices of this study (

Table 3. The semantic differential scale.

Symbol	Variable	Index	Definition
M1	Timbre	Low frequency-High frequency	Frequency performance of the sound
M2	Tone	Heavy-Sharp	Thickness and sharpness of the sound
M3	Loudness	Low-High	Subjective loudness of the sound
M4	Sound source distance	Close-Far	Distance between sound source and interviewee
M5	Sound source number	Small-Large	Number of sound sources
M6	Coverage	Small-Wide	Influence and coverage of the sound
M7	Clarity	Vague-Clear	Vagueness and clarity of the sound
M8	Order	Disordered-Ordered	Disorder and order of the sound
M9	Comfort level	Uncomfortable-Comfortable	Comfort brought by the sound
M10	Sense of belonging	Strange-Familiar	Strangeness and familiarity brought by the sound
M11	Sense of security	Dangerous-Safe	Danger and safety brought by the sound

). In this study, there were five evaluation scale levels for each index; that is, there were five interval values between each adjective combination, and the value of 3 was the symmetrical midpoint. Based on this, the residents needed to score different sound types. A higher score of each evaluation item indicates a tendency toward the right adjective, a lower score of each evaluation item indicates a tendency toward the left adjective, and a score around 3 indicates a score with little to no tendency toward one or the other. To better highlight the differences between various soundscape types without making the conclusions too detailed, the SD score of this study was aimed at the urban park as a whole.

2.3.2. Importance-Performance Analysis (IPA) Model

The importance-performance analysis (IPA) model was used to analyze and measure the residents’ comprehensive perceptions of urban park soundscapes [59,60]. This model was divided into four quadrants, with importance and performance as the axes. Through quadrant analysis of the importance and performance, the results of the residents’ comprehensive perceptions were divided into four areas. The first quadrant in the upper right corner is the advantage area; residents believe that the factors in this area are very important, and the soundscape performances of these factors in urban parks are very good. The second quadrant in the upper left corner is the improvement area; residents believe that factors in this area are very important, but the soundscape performances are not ideal. For factors in this quadrant, the park should pay more attention to planning and managing them, especially focusing on repairs and improvements. The third quadrant in the lower left corner is the opportunity area; residents believe that the included factors are not important, and the soundscape performances are relatively poor. However, this does not imply that the factors in this area can be ignored in the soundscape construction of urban parks. Instead, special attention should be paid to analyze the causes of these factors to find methods for improving the residents’ satisfaction. The fourth quadrant in the lower right corner is the maintenance area; residents believe that the importance rating of factors in this area is relatively low, but their soundscapes perform well. Therefore, effectively utilizing these factors may play a positive role in improving the soundscapes of urban parks. In the implementation of the IPA model, this study used a five-point Likert scale to measure the residents’ perceptions of the importance and performance of soundscapes (Figure 6).

2.3.3. Multinomial Logit Model

Calculations and analyses were conducted by constructing a disordered multinomial logit model and comparing the disordered choices of residents. Multinomial logit models were applied to the analysis of the residents’ preferences for the three types of soundscapes (Y). In this case, Y is a function of the individual characteristics of the residents’ choices. Specifically, multinomial logit regression involves more than two types of nominal response variables, and the response variables with (j + 1) types produce j formulas, each of which is a binary logit regression process compared with the control group. In the multinomial logit regression, the above j binary logit regressions were estimated simultaneously. If the last group is set as the control group, the other groups are experimental groups, P (y_i = j) represents the probability of the experimental group, P (y_i = 0) represents the probability of the control group, and β refers to the log odds of the experimental group for the control group, then the form of the multinomial logit model can be set as follows:

$$P\left(y_i=j\right)=e^{x_i{\beta }_j}/(1+{{\displaystyle \sum }}_{k=1}^j{e^{x_i{\beta }_k}}^{\prime })$$

(1)

$$P\left(y_i=0\right)=1/(1+{{\displaystyle \sum }}_{k=1}^j{e^{x_i{\beta }_k}}^{\prime })$$

(2)

When estimating each model, the standardized coefficient of the control group was zero. This is because the sum of the probabilities for all options must be uniform. Based on the logit model, the relative probability for the reference group can be obtained as:

$$P\left(y_i=j\right)/P\left(y_i=0\right)=e^{x_i{\beta }_j},$$

(3)

$$ln\left(P\left(y_i=j\right)/P\left(y_i=0\right)\right)=x_i{\beta }_j,$$

(4)

Corresponding calculation methods also exist for estimating the parameters of disordered multinomial logit models. For the i^th resident, if the j^th scheme is selected, d_ij = 1; if the j^th scheme is not selected, then let d_ij = 0. At the same time, for the i^th resident, only one of the (j + 1) alternatives can be selected, that is, only one d_ij = 1 exists. Therefore, the joint probability function of y_ij (i = 1, 2,..., n; j = 0, 1, 2,..., m) can be written, obtaining the likelihood function, and finally obtaining the log-likelihood function as:

$$lnL={{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^m{d}_{ij}lnP\left(y_i=j\right),$$

(5)

It should be noted that the non-correlation test of the original data was carried out before the disordered multinomial logit models were used for the estimation. Based on this, it can be determined whether disordered multinomial logit models can be used for parameter estimations, and whether it is meaningful to study this problem.

3. Results

3.1. Demographic Characteristics of the Respondents

Regarding the individual characteristics (shown in

Table 4. Descriptive statistics of the samples.

Variable	Frequency (%) (n = 749)	Variable	Frequency (%) (n = 749)
(Y) Preferred soundscape types		(X4) Marital status
Human sounds	196 (26.17)	Unmarried	332 (44.33)
Artificial sounds	286 (38.18)	Married	417 (55.67)
Natural sounds	267 (35.65)	(X5) Family size
(X1) Gender		1 person	62 (8.28)
Men	376 (50.2)	2 persons	102 (13.62)
Women	373 (49.8)	3 persons	266 (35.51)
(X2) Age		4 persons	202 (26.97)
Age 17 and under	12 (1.6)	5 persons or more	117 (15.62)
18–29 years old	324 (43.26)	(X6) Health status
30–39 years old	193 (25.77)	Very good	10 (1.34)
40–49 years old	92 (12.28)	Relatively good	26 (3.47)
50–59 years old	70 (9.35)	Neutral	186 (24.83)
Age 60 and over	58 (7.74)	Relatively poor	349 (46.6)
(X3) Residential area		Very poor	178 (23.77)
Within the Inner Loop	18 (2.4)	(X7) Residence status
Inner Loop-Middle Loop	80 (10.68)	Native born	236 (31.51)
Middle Loop-Outer Loop	168 (22.43)	First-generation immigrants	100 (13.35)
Outer Loop-City boundary	340 (45.39)	Second-generation immigrants	177 (23.63)
Outside the city border	143 (19.09)	Foreign-registered residence	236 (31.51)

), the proportions of male and female respondents differed slightly but were generally balanced. More than 70% of respondents were under 40 years old, and the proportion of married respondents was about 10% higher than those unmarried people. The overall family size of the respondents was three or four, and the proportion of people living alone was very small. Most of the respondents’ physical conditions were at an average level. Furthermore, nearly 70% of the respondents lived between the Middle Loop and the city boundary, which was consistent with the expected number and the population density of residents in Shanghai. In the aspect of residence status, except that the proportion of the first-generation immigrants was slightly low, the proportion of residents of native born, second-generation immigrants, and foreign-registered residence was relatively close. This sample distribution result obtained by random sampling revealed the current structural proportion of the residence status in Shanghai, which shows that the current composition of Shanghai residents is quite diversified, which is a meaningful survey result.

3.2. Soundscape Characteristics Results

The soundscape experience results show that the sound level changes in different periods are closely related to its own functions and the residents’ activities in the park. The formal survey was conducted in May 2021, which included the International Labor Day holiday. Thus, the characteristics of the Century Park soundscape were more obvious, and the results yielded were more consistent. The specific results are shown in Figure 7.

Specifically, the period from 8:00 to 10:00 is the morning peak time in Shanghai, so the sound level of artificial sounds continues to improve. During the period from 10:00 to 11:00, most residents who need to go to work have basically arrived at the workplace, while other residents begin to come to Century Park for recreational activities. The entertainment projects in the park are gradually enriched and the flow of people is increasing. At this time, human sounds such as entertainment activity and playing become the main sound type of Century Park. At the same time, with the increase in residents in Century Park and the opening of businesses and entertainment facilities, human sounds of vendors and artificial sounds of music also begin to appear in the park. During the period from 11:00 to 12:00, the sound level of human sounds reached 72.97 db, while the sound level of natural sounds also increased significantly but did not exceed 63 dB. During the period from 12:00 to 13:00, some residents who came to Century Park in the morning began to leave, but a large number of residents still carried out catering, lunch breaks, and other activities near the park, so the sound level generally decreased, in particular, the human sounds decreased significantly to 57.65 db. From 13:00 to 16:00, the number of residents in the park began to increase again, and various entertainment activities appeared again. Human sounds of playing and chatting appeared all over the park, and the sound level generally improved. The period from 16:00 to 18:00 is the evening peak time in Shanghai, and the artificial sounds are constantly improving. At about 18:00, Century Park will close on time, the entertainment facilities in the park are also closed, and the residents gradually leave the park. Therefore, the sound levels of human sounds and natural sounds decline significantly.

In addition, according to the environmental noise limit in the national standard of the People’s Republic of China for acoustic environmental quality standard (Standard number: GB 3096-2008), combined with the sound level measurement results in Century Park, this study defined the measurement results with a sound level above 65 dB as “noisy”. Referring to this standard, it can be seen that the artificial sounds in Century Park are noisy during the morning and evening peak, the human sounds are noisy during 10:00 to 12:00 in the morning and 13:00 to 17:00 in the afternoon, the time periods being the time when residents gather in the park. In contrast, natural sounds are not noisy at all times of the day.

3.3. Analysis of Century Park Soundscape Characteristics

After conducting the calculations, SD scoring sheets for the three soundscape types were obtained (

Table 5. Scale of the semantic differences for different sound types.

Symbol	Variable	Index	Human Sounds	Artificial Sounds	Natural Sounds
M1	Timbre	Low frequency-High frequency	4.037	4.402	3.663
M2	Tone	Heavy-Sharp	3.252	4.478	2.500
M3	Loudness	Low-High	3.703	4.530	2.310
M4	Sound source distance	Close-Far	2.972	4.159	3.333
M5	Sound source number	Small-Large	4.089	3.590	2.754
M6	Coverage	Small-Wide	4.187	3.821	3.341
M7	Clarity	Vague-Clear	2.736	3.916	2.385
M8	Order	Disordered-Ordered	2.528	2.116	3.464
M9	Comfort levels	Uncomfortable-Comfortable	2.484	2.279	3.722
M10	Sense of belonging	Strange-Familiar	3.285	2.339	4.246
M11	Sense of security	Dangerous-Safe	4.118	2.522	3.631

). Based on the SD score, the corresponding SD evaluation curve (Figure 8) was drawn to show the characteristic differences more intuitively between different soundscape types to better reflect the residents’ perceptions of the advantages and disadvantages of different soundscape types.

In general, the differences in the perceptions of various soundscapes were large regarding loudness, tone, and sense of belonging, whereas the differences regarding timbre, coverage, and sound source distance were relatively small. Specifically, these results can be summarized into the following three aspects: (1) The advantages of natural sounds are that the tone is relatively heavy, the loudness is relatively low, and the sound is relatively ordered, this may be one of the reasons why residents have a more comfortable perception of natural sounds. However, natural sounds also have certain disadvantages. For example, they have a small number of sound sources, smaller coverage, and lower clarity of soundscapes compared to artificial and human sounds. (2) For artificial sounds, although their sound source distance is a bit longer, their tone is relatively sharp, and the loudness is relatively strong. Similarly, the clarity of artificial sounds is the highest and the order degree is the lowest, which causes these sounds to bring about strongly uncomfortable, strange, and dangerous feelings in urban parks. (3) The main characteristics of human sounds are a large number of sound sources and the largest coverage. Human sounds are moderate in timbre, tone, loudness, and order degree, and they are relatively similar to the soundscapes present in the residents’ daily lives. Furthermore, the sense of security brought to residents was the strongest from human sounds, although the appearance of human sounds at a high frequency, number, and range will also cause a less comfortable perception.

3.4. Perceptions of Century Park Soundscapes

In this study, the residents’ subjective perceptions of the soundscapes were further discussed based on their evaluations of the degree of favorability, importance, and subjective loudness of the soundscapes. The results are listed in

Table 6. The evaluation results of the degree of favorability, importance, and subjective loudness of soundscapes.

Sound Types	Favorability Degree					Importance Degree					Subjective Loudness Degree
	(1)	(2)	(3)	(4)	AVG	(1)	(2)	(3)	(4)	AVG	(1)	(2)	(3)	(4)	AVG
Human sounds
Entertainment activity	4.11	3.23	4.27	3.26	3.72	3.60	4.53	2.37	3.42	3.48	4.23	4.15	4.33	4.20	4.23
Playing	4.03	3.24	4.28	2.49	3.51	2.62	3.91	2.45	2.48	2.87	4.40	3.01	4.72	4.05	4.04
Vendors	3.19	3.44	3.38	3.85	3.47	4.62	4.06	3.26	3.71	3.91	4.27	4.62	4.54	4.09	4.38
Chatting	3.15	4.55	3.55	2.64	3.47	2.42	3.70	4.47	4.59	3.79	4.62	3.12	4.90	4.50	4.28
Average	3.62	3.62	3.87	3.06	-	3.32	4.05	3.14	3.55	-	4.38	3.72	4.62	4.21	-
Artificial sounds
Traffic	3.81	4.19	4.44	3.15	3.89	4.12	2.35	2.53	3.26	3.06	4.86	4.57	4.35	4.66	4.61
Entertainment equipment	3.30	3.61	3.03	3.03	3.24	3.21	4.14	3.22	3.08	3.41	4.51	4.44	4.50	4.95	4.60
Music	4.69	4.35	4.81	3.63	4.37	2.36	2.32	2.61	2.67	2.49	2.41	3.04	4.23	4.46	3.54
Construction	2.43	2.91	2.27	3.61	2.81	3.65	3.64	2.84	3.20	3.33	4.94	4.84	4.89	4.14	4.70
Average	3.56	3.77	3.64	3.36	-	3.34	3.12	2.80	3.05	-	4.18	4.22	4.49	4.55	-
Natural sounds
Wind	4.11	4.51	4.10	4.58	4.32	3.07	4.07	4.11	3.44	3.67	3.15	4.28	3.93	2.89	3.56
Birdsong	4.46	4.16	3.98	4.51	4.28	4.07	4.56	4.01	3.55	4.05	4.35	4.64	3.62	3.94	4.14
Leaves	3.76	3.81	4.37	3.86	3.95	3.40	3.05	4.84	3.40	3.67	4.18	3.79	4.20	2.52	3.67
Water	3.87	4.15	4.27	4.39	4.17	3.34	4.71	3.30	4.37	3.93	3.26	3.82	3.55	3.13	3.44
Average	4.05	4.16	4.18	4.33	-	3.47	4.09	4.06	3.69	-	3.73	4.13	3.83	3.12	-
Total average	3.74	3.85	3.90	3.58	-	3.37	3.75	3.33	3.43	-	4.10	4.03	4.31	3.96	-

Note: (1)–(4) represent native-born residents, first-generation immigrants, second-generation immigrants, and foreign-registered residents, respectively.

. In general, the residents found natural sounds to be most favorable. Specifically, the residents found the sounds of music (4.37), wind (4.32), and birdsong (4.28) to be the most favorable. This result reflects the general desire of residents to be close to nature. In contrast, construction sound (2.81) was the soundscape most disliked by residents. Such sounds cover the natural sounds present in the park such as birdsong, and they reduce the residents’ perceptions of the soundscapes. Based on the interviews conducted in this survey, residents are motivated to use urban parks like Century Park as important places to spend their leisure time, return to nature, and relax their bodies and minds. The residents responded that they paid great attention to the healing effects of urban parks on their souls, which were exhausted from a fast-paced urban lifestyle. Using birdsong as an example, it has been pointed out that bird chirping can cause a positive emotional response. Therefore, this study considered that the negative effect caused by unnatural soundscapes on natural sounds may be one of the reasons why the residents’ overall level of soundscape preference in urban parks was low.

Regarding the evaluations of subjective loudness, the soundscapes with higher subjective loudness scores were mainly those with artificial sounds and human sounds, indicating that the volume of these sounds is higher and has a wider spread. The results of this survey are consistent with reality. As the largest urban park in Shanghai, Century Park attracts many residents as visitors. Various kinds of artificial and human sounds appear in the park; thus, the soundscape loudness of the park is relatively large. However, with the exception that the loudness of music was accepted and understood, residents evaluated most of the artificial and human sounds with low favorability; these results show that artificial and human sounds have some negative impacts on the park-touring experiences of residents. In addition, although the residents most favored natural sounds, they also thought that the subjective loudness of natural sounds was low, which indicates that the characteristics of natural sounds are not obvious in Century Park. Furthermore, the interviews showed that residents tended to pay more attention to visual experiences when they were in the park, and many people were not very familiar with the concept of soundscapes. This study shows that the low familiarity with soundscapes is the reason for the low subjective perception of residents, which makes the existing soundscape unable to meet the psychological needs of residents.

There was a certain degree of correlation between soundscape favorability and subjective loudness, but the relationship differed depending on the types of soundscapes and the residents’ identities (

Table 7. The evaluation results of the degree of correlation between soundscapes.

	Residence Types	Native-Born	First-Generation Immigrants	Second-Generation Immigrants	Foreign Registered Residence
Sound Types
Human sounds		−0.48	−0.41	−0.36	−0.33
Artificial sounds		−0.82	−0.71	−0.95	−0.89
Natural sounds		0.31	0.49	0.51	0.63
Total		−0.64	−0.30	−0.51	−0.64

). The correlation coefficient is a statistical index used to reflect the close degree of correlation between variables. Due to the different research objects, the correlation coefficient can be defined in many ways. In this study, the CORREL function of Excel software was used to calculate the correlation coefficient between the favorability and subjective loudness to determine the degree of correlation between the two. The results showed that there was a negative correlation between the favorability and subjective loudness of human and artificial sounds, whereas there was a positive correlation between those of natural sounds. The negative correlation was most obvious for artificial sounds and second-generation immigrants (R = −0.95). Some types of soundscapes were too loud and noisy, which conflicted with the relaxing park environment. The above results show that there is an inseparable relationship between soundscape favorability and the degree of subjective loudness of urban parks, which demonstrates that the soundscape environments of urban parks are a unified whole. However, it is necessary to point out that there were differences in the correlation between the different residents’ perceptions of soundscape favorability and subjective loudness, indicating that the relationship between the two is not linear and may also be affected by other factors such as the residents’ individual preferences or routines.

3.5. Importance-Performance Analysis (IPA) of Soundscapes

The residents’ perceptions and preferences for urban park soundscapes were attributed to how they evaluated the importance and performance of urban park soundscapes. The constructed IPA grids reflect the importance of soundscape elements and their average performance, showing the differences between different observation points. The distribution of the point positions representing the importance and average performance of the 12 sound types is shown in the IPA grids in Figure 9.

Specifically, all of the residents’ perceptions of birdsong and wind sounds were in the first IPA quadrant, indicating that these two sounds are very important and highly favored by the residents; these perceptions should continue. Generally speaking, most of the IPA perceptions of natural sounds were located in the dominant area, indicating that residents had a higher degree of favorability for all kinds of natural sounds. However, some natural sounds were perceived to have insufficient importance such as the perceptions of wind sound by native-born residents and the perceptions of leaf sounds by first-generation immigrants and residents with foreign-registered residences. Compared with the importance perceptions of natural sounds, those of most artificial sounds were low, and the IPA perception was mostly located in the third and fourth quadrants. The IPA perceptions of music among all residents were in the fourth quadrant, indicating that residents generally tend to have a higher degree of favorability for music, but they have insufficient cognition of its importance, which is a common problem. According to the interviews, it can be seen that most residents thought music was the icing on the cake in the urban park soundscapes. However, music was considered “dispensable” compared to sound types with higher importance such as birdsong. Therefore, urban parks should continue to design and incorporate music and similar soundscapes in the park while also raising public awareness of the importance of such soundscapes to help them continue to play a positive role. The IPA perceptions of vendor sounds by native-born residents and first-generation immigrants, and the perceptions of chatting sounds by second-generation immigrants and residents with foreign-registered residences were all located in the second quadrant. This shows that residents think the human sounds of vendors and chatting are very important, but their performances in urban parks are not ideal and they do not meet the residents’ expectations. In the future, more attention should be paid to the restoration and improvement of such soundscapes to encourage more favorable evaluations by residents and to encourage these soundscapes to gradually meet the residents’ expectations. The residents’ IPA perceptions of artificial sounds such as entertainment equipment and construction sounds were mostly located in the third quadrant and its adjacent areas. Residents thought that the importance of such soundscapes was low and that their performance in urban parks was relatively poor. However, this does not mean that the soundscape planning of urban parks can ignore the soundscape types located in the third quadrant; instead, reasons for the residents perceiving these soundscapes with low favorability should be carefully analyzed to ultimately develop ways to improve these perceptions.

The above results indicate that there is a certain correspondence between the degree of favorability and the perceived importance of urban park soundscapes. The residents’ IPA perception results for all soundscape types also reflected certain distribution laws, and the differences in the residents’ identities also had an impact on the perception results. In general, the IPA results fully demonstrate the characteristics of all types of soundscape environments in urban parks.

3.6. Soundscape Preferences Based on Individual Differences

To explore the influence of individual differences on the residents’ preferences for urban park soundscapes, this study considered the residents’ preference for three types of soundscapes (Y) as the explained variable of the model. Before performing parameter estimation, a test for independence was required for the original data. The original hypothesis of this test is that the regression coefficient of the model is 0, so if the independence test results show that p < 0.05, it means that the original hypothesis can be rejected; that is, the model is effective. The calculation results showed that in the interpretation results of all variables shown in

Table 8. Multinomial logit model for the residents’ individual characteristics.

Index	Multinomial Logit Model (N = 749)				Optimal Preference
	$\ln \frac{\mathrm{P}\left({\mathrm{y}}_i=1\right)}{\mathrm{P}\left({\mathrm{y}}_i=3\right)}$		$\ln \frac{\mathrm{P}\left({\mathrm{y}}_i=2\right)}{\mathrm{P}\left({\mathrm{y}}_i=3\right)}$
	B	Exp(B)	B	Exp(B)
(C)	−0.87	-	−0.68	-	-
Gender = Men	−0.2	0.82	−0.06	0.94	Natural sounds
Age = Age 17 and under	−1.25	2.38	−1.26	4.53	Artificial sounds
Age = 18–29	0.87	1.77	1.51 ***	0.78	Human sounds
Age = 30–39	0.65	1.92	0.38	1.46	Human sounds
Age = 40–49	0.74	2.10	0.41	1.50	Human sounds
Age = 50–59	0.57	0.29	−0.25	0.28	Natural sounds
Residential area = Within the Inner Loop	0.51	1.66	0.98	2.67	Artificial sounds
Residential area = Inner Loop–Middle Loop	0.13	1.14	0.16	1.18	Artificial sounds
Residential area = Middle Loop–Outer Loop	−0.69 **	0.50	−0.13	0.88	Natural sounds
Residential area = Outer Loop–City boundary	−0.15	0.86	−0.31	0.73	Natural sounds
Marital status = Unmarried	0.51	1.67	−0.01	0.99	Human sounds
Family size = 1 person	0.99 **	2.70	1.06 ***	2.89	Artificial sounds
Family size = 2 people	1.23 ***	3.43	1.27 ***	3.54	Artificial sounds
Family size = 3 people	0.98 ***	2.67	1.07 ***	2.92	Artificial sounds
Family size = 4 people	0.26	1.29	0.31	1.36	Artificial sounds
Health status = Relatively good	−0.94 *	0.39	−1.4 **	0.25	Natural sounds
Health status = Average	−0.57 **	0.56	−0.07	0.93	Natural sounds
Health status = Relatively poor	0.12	1.12	−0.11	0.89	Human sounds
Residence status = Native-born	−0.71 *	0.49	−0.66 *	0.52	Natural sounds
Residence status = First-generation immigrant	−0.32	0.73	−0.9 ***	0.41	Natural sounds
Residence status = Second-generation immigrant	−1.43 ***	0.24	−1.48 ***	0.23	Natural sounds
Residence status = Foreign-registered residence	0.07	1.08	0.64 *	1.90	Artificial sounds

* represents 90% statistical significance, ** represents 95% statistical significance, *** represents 99% statistical significance.

, χ = 102.354 and p < 0.050, indicating that the original hypothesis could be rejected, and the model can proceed; that is, the independent variables included when the model was under construction were effective. This model construction was therefore meaningful, and the model could explain the research problems well. In conclusion, this study met the requirements of using a logit model to estimate the studied parameters. Based on this, the study used the maximum likelihood method to estimate the parameters of Y, and variables with the parameter estimation values of zero were eliminated based on the Wald test results to obtain the final results (Table 7). In this process, schemes 1 (human sounds) and 2 (artificial sounds) were set as the experimental groups and scheme 3 (natural sounds) was set as the control group.

Overall, the residents’ preferences for different soundscape types varied widely. Older residents (age >50 years) preferred natural sounds, which is consistent with the living habits and needs of the elderly, who have more leisure time and like to be close to nature. Additionally, unmarried residents had a higher preference for human sounds. These residents have higher physical activity levels, more flexible time arrangements, and greater participation in various recreational activities in the park. Therefore, they had a relatively high preference for sounds such as those from entertainment, playing, and chatting. Residents living outside the Middle Loop area preferred natural sounds, whereas those in this area preferred artificial sounds. According to the interviews, the population density within the Middle Loop of Shanghai was higher and the built-up area was larger compared to the other areas. Residents who lived outside of this area had a higher demand for the natural environment. They were more eager to throw off the shackles of urban life, stay away from built-up areas with noisy soundscapes, and enjoy the freedoms and comforts brought by areas outside the Middle Loop. In terms of health status, residents in good health preferred natural sounds, and this preference corresponds to their healthy habit of being close to nature. Furthermore, family size did not have a differential impact on the residents’ preferences.

The conclusions regarding the residence status require further discussion. The results showed that residents with Shanghai registered residences (i.e., native-born residents and first- and second-generation immigrants) preferred natural sounds. Some studies have pointed out that residents with an understanding of the location of urban parks is a prerequisite for their use [61]. The interview results showed that residents with registered residences in Shanghai seemed to have a better understanding of the existence, function, and use of Shanghai city parks than the other residents. Based on these cognitive experiences, these Shanghai-registered residents tended to have a stronger sense of identity with the city. In recent years, Shanghai has put forward goals for achieving a higher quality of living environments, and the residents’ requirements and expectations for the city’s natural ecological constructions have increased accordingly. Therefore, residents are paying more attention to and prefer natural sounds. In contrast, most of the activities of people in Shanghai with foreign-registered residences are simple working, commuting, and traveling in Shanghai. Thus, they do not have a high demand for the natural environment of the city, and they instead pay more attention to the various urban socioeconomic functions of Shanghai as a megacity and regional economic center. Therefore, these residents showed a high preference for artificial sounds.

4. Discussion

Based on the research conclusions, this study further discussed the effects of soundscapes on the residents’ experiences in Century Park from the following four aspects:

Through comprehensive and systematic analysis, this study analyzed the characteristics of various types of urban park sounds, thus provides a reference and basis for the development of urban parks that have more practical value and natural ecological benefits. The soundscape experience results show that the sound level changes in different periods are closely related to its own functions and the residents’ activities in the park. Specifically, the measurement results of sound level show great differences in different time periods of a day, and the change in the sound level is related to the increase or decrease in the number of people in the park in different time periods. Overall, the periods with higher sound levels of artificial sounds often overlap with the morning and evening peaks in Shanghai, while the human sounds are noisy in the time period when residents gather in the park. The soundscape survey showed that the perceived differences in the soundscapes regarding loudness, tone, and sense of belonging were large, whereas the differences regarding timbre, coverage, and sound source distance were relatively small. Specifically, the loudness of natural sounds was relatively low and ordered, which provided residents with soundscape experiences that were familiar in terms of their sense of belonging and comfort. However, natural sounds had a smaller number of sound sources, and their coverage was relatively small. Thus, to expand the coverage and influence of natural sounds, it is necessary to increase the number of sound sources through actions such as increasing the vegetation coverage and protecting the fauna and flora in the park. In contrast, the tone of artificial sounds was sharp, the volume was relatively high, and the order degree was very low; these factors caused artificial sounds to bring residents uncomfortable, strange, and dangerous feelings, which reduced their overall experiences of urban park soundscapes. Therefore, the next step regarding artificial sounds is to manage and limit some of them and focus on concentrated problems. In contrast to the other two types of sound, human sounds had more sound sources and the widest coverage. The soundscape characteristics of human sounds had no outstanding advantages or disadvantages, but they were an indispensable type of soundscape in the residents’ daily lives. For example, human sounds provided residents with the greatest sense of security compared to the other types of sound, indicating that urban parks need to provide necessary social service functions in addition to natural functions. Both human and natural sounds played an important role in the residents’ soundscape experiences, and this result corresponded with those of previous research [62,63]. Furthermore, due to the lack of soundscape-related knowledge and experiences in the management department of Century Park (as determined based on the interviews), it is difficult to balance the relationship between natural sounds, human sounds, and artificial sounds, which results in interactions among various soundscapes in the urban park, providing residents with noisy and chaotic soundscape experiences. These are problems that must be solved urgently.

Regarding the perceptions of soundscapes in Century Park, all types of residents had a high preference for natural sounds; therefore, it is vital for Century Park to protect and cultivate its natural sounds. In fact, there are many sources of water sounds in Century Park such as fountains and streams, but there is a small number of sound sources that residents can experience. The reason is that the role of water sounds in the park cannot be brought into full effect because of the limited equipment and opening hours of the park as well as the coverage of other types of sounds in the park. Against the background of traffic and other noises, the positive impact of water sounds on residents was quite limited. Therefore, the opening hours of fountains and other facilities in the park should be appropriately increased, and the water sound in the park should be further increased by adding small waterscape facilities. For example, artificial streams and waterfalls could be added to the park, where the park could refer to the construction methods of Japanese garden waterscapes, and it could incorporate additional facilities such as waterwheels beside the streams to ensure that the park has rich water sounds for residents to experience when the fountains are closed. Similarly, due to the high green coverage of Century Park, birdsong has become a major feature of natural sounds in the park. However, due to the high volume of artificial and human sounds in the park, birdsong and other natural sounds were not obvious to the residents. To enrich the natural sounds in Century Park, different kinds of plants could be added alongside the original plants to attract more birds to forage and inhabit the area. Additionally, bird activities can produce pleasant sounds of rustling leaves, which can further increase the proportion of natural sounds in the park.

In addition, considering that the urban park is in the built-up area of the city and its comprehensive functional characteristics, noise caused by artificial sounds such as traffic, entertainment equipment, and construction is inevitable. However, it can be alleviated by covering and absorbing these sounds to reduce the impact of noise on the overall soundscapes of Century Park. For example, trees and plants could be planted in the entrance area of the park as a noise barrier, which would absorb or reflect sounds by relying on the resonance generated by leaves and trunks to block and reduce noise. The noises caused by humans such as entertainment activity, playing, vendors, and chatting in the park can be alleviated by formulating the corresponding standards and regulations. For example, for entertainment activities that require music played on large-scale stereo equipment, the volume should be specified within a certain standard. As another example, Century Park should provide real-time statistics on the clustering of residents in different areas of the park and limit the foot traffic in some key areas to alleviate the noise caused by excessive foot traffic. The results of this study showed that there was an inseparable relationship between the favorability and subjective loudness of urban park soundscapes, and that the soundscape environments of urban parks were a unified whole. Therefore, the main sound sources causing noise must be controlled so that the various types of soundscapes in the park will not interfere with each other and will instead be coordinated.

The results of the importance performance analysis (IPA) demonstrated that there were certain corresponding relationships between the favorability and perceived importance of urban park soundscapes. The IPA perception results of the residents for various soundscapes also reflected certain distribution laws, and the difference in the residents’ identities had an impact on their perceptions. First, the IPA perceptions of most natural sounds, especially birdsong and wind sounds, were in the first quadrant, which showed that the residents viewed natural sounds as positively favorable and important. However, some natural sounds lacked importance; the park should pay more attention to these sounds. Second, the residents’ perceptions of the importance of the most artificial sounds were low. Taking music as an example, the residents viewed music with a high degree of favorability, but a relatively low importance. In essence, this showed that in the perceptions of urban park soundscapes, the residents’ perceptions of the so-called “social soundscapes” produced by human society were often lower than those of the sounds existing in the natural environment. Therefore, it is important to design and plan social soundscapes involving music in urban parks. The goal of relevant work should be to encourage social soundscapes to play a more positive role to enhance the residents’ understanding and attention of such sounds. Third, the residents believed that human sounds such as vendors and chatting were very important, but the corresponding degree of favorability was low; therefore, more focus needs to be placed on the repair and improvement of this kind of soundscape. Finally, the residents’ evaluations of the importance and favorability of artificial sounds such as entertainment equipment and construction sounds were relatively low, indicating that such soundscapes do not play a positive role in urban parks. This study shows that relevant departments need to strengthen the operation and management of artificial sounds and their facilities to ensure that residents are provided with positive soundscape experiences. However, the management department of Century Park does not maintain or actively seek to improve the soundscape of the park. Therefore, the park management department should take appropriate responsibility, enhance its professional quality, and carefully analyze and solve the problems related to artificial sounds to ultimately improve how residents view the soundscapes in the park.

Residents have different soundscape perception experiences due to their own natures and states. Under different individual characteristics, the residents’ preferences for different soundscape types were quite different. In other words, for the improvement and optimization of Century Park soundscapes, not only should the relevant managers strengthen their management, but more attention should also be paid to the differential preferences and demands of the residents. For example, full consideration should be given to the living habits of older residents who have more leisure time and like to be close to nature; this can be partially achieved by improving the level of natural sounds in the park. As another example, the stronger physical activity levels and flexible time arrangements of unmarried residents should be taken into account when considering maintaining a reasonable proportion of human sounds in the park. Furthermore, it is necessary to fully understand that residents outside the Middle Loop area prefer the natural sounds of the park due to the characteristics of their residential areas. Overall, the individual differences of different types of residents should be considered when building and improving the park’s soundscapes. In addition, the results showed that native-born residents and first- and second-generation immigrants preferred natural sounds; that is, residents with registered residences in Shanghai seemed to know more about the existence, function, and use of the city’s urban parks than the other residents. In contrast, the activities of people in Shanghai with foreign-registered residences mainly include working, commuting, and traveling. The demands for the natural environment in Shanghai are not high, and they pay more attention to the social and economic functions of Shanghai as a megacity and regional economic center; thus, they have a higher preference for artificial sounds.

In addition, this study has many limitations that need to be further addressed in the future. First, there is a lack of comparison between multiple parks. Due to limited time and energy, this study only considered Shanghai Century Park as the research area. Century Park is a large, comprehensive urban park that combines ecology with entertainment, and its research results are representative and typical. However, the results of this study cannot accurately reflect the actual situation of all urban park soundscapes. In future research, it is necessary to investigate more parks, comprehensively grasp more elements of urban park soundscapes, and obtain more scientific strategies for improving and optimizing such soundscapes so that the research conclusions can be more relevant and valuable. Second, this study on Century Park was only conducted in the spring, therefore causing the study design to lack time continuity. Seasonal changes may cause changes in the natural environment, user types, and user perceptions and behaviors in urban parks, thus affecting the results of the study. In the future, a comparative study of urban park soundscapes across all four seasons needs to be considered to draw more comprehensive conclusions. Finally, to improve and optimize urban park soundscapes, the relationships between soundscapes and other perceived experiences should also be considered and their interrelationships should be further analyzed. Based on this, it can be better ensured that the research conclusions can provide meaningful guidance for the comprehensive and coordinated development of urban parks.

5. Conclusions

This study considered three typical types of urban park soundscapes: natural sounds, artificial sounds, and human sounds. Then, a comprehensive analysis using the semantic differential method, the importance-performance analysis (IPA) model, and multinomial logit models was conducted. Shanghai Century Park was chosen as the research area for conducting quantitative and qualitative research on urban park soundscapes. This study comprehensively discussed the residents’ perceptions and differential preferences for urban park soundscapes under different individual characteristics. The results showed that the artificial sounds in Century Park were noisy during the morning and evening peak, while the human sounds were noisy when residents gathered in the park. There was a close relationship between the favorability and subjective loudness of urban park soundscapes, and the residents’ perceptions of urban park soundscapes substantially differed in terms of loudness, tone, and sense of belonging, whereas the differences in the perception of timbre, coverage, and sound source distance were relatively small. It is necessary to enhance the role of natural sounds as well as to control the main noise-producing sound sources so that various sounds in the park will not interfere with each other; the optimization of soundscapes should also focus on the residents’ different soundscape preferences due to their individual characteristics. This study demonstrates that the participation and feedback behaviors of residents are very important in the process of improving and optimizing urban park soundscapes. The ultimate purpose of soundscape design and its functions in urban parks is to meet the changing demand for the natural environment and the corresponding perceptions of urban residents. Only by fully understanding the residents’ perceptions and individual preferences can the most favored urban park soundscape be formed to achieve the effective coordination and realize the value of various urban park soundscapes. At the same time, improving the residents’ cognition and understanding of Century Park soundscapes can greatly reduce the negative experiences caused by a lack of understanding of soundscapes.