The Impact of Visual, Thermal, and Acoustic Environments in Urban Public Spaces in Cold Regions on the Psychological Restoration of the Elderly

Abstract

Optimizing the visual, thermal, and acoustic environments of urban public spaces in severely cold regions can significantly enhance the psychological restoration of the elderly, addressing the increasing mental health demands in an aging society. Despite its importance, the mechanisms, strategies, and seasonal effects of various environmental variables on psychological restoration remain inadequately studied. This research uses Harbin as a case study, employing field surveys and tests to systematically examine the elderly’s psychological restoration across different seasons. By integrating environmental stimulus variables with a psychological restoration evaluation model, the study investigates the impact of urban public spaces on mental health. The key findings are: (1) The spring environment of urban public spaces has the most significant positive effect on psychological restoration, with an effect size of η² = 0.360. (2) A significant correlation exists between environmental variables and psychological restoration year-round, with the panoramic green view index in winter showing the highest positive impact (correlation coefficient = 0.301, p < 0.01). (3) The influence of environmental stimulus variables on psychological restoration varies notably across seasons; the acoustic environment in spring contributes most significantly, with an R² = 17.03%, while visual factors dominate in winter and summer. (4) Conditional probability analysis reveals the effects of various environmental variables on psychological restoration, proposing season-specific environment optimization strategies. Based on these findings, the paper presents a model for optimizing urban public space environments in severely cold regions, aiming to maximize elderly psychological restoration by tailoring environmental stimulus variables to their mental health needs.

1. Introduction

1.1. Research Background and Origin

With the acceleration of globalization, urbanization and population aging have become two prominent trends in current societal development. These trends have not only profoundly altered the global demographic structure but have also posed new challenges to the quality of urban life [1,2]. The rapid expansion of cities has gradually distanced many elderly residents from natural environments, adversely affecting their mental health [3,4]. Specifically, a range of issues in urban environments, such as the urban heat island effect [5,6], air pollution [7,8], and noise pollution [9,10], have significantly increased stress, anxiety, and other negative psychological states among the elderly population. It is estimated that by 2050, over one-sixth of the global population will be 65 years old or older, a trend that will exacerbate mental health issues among the elderly [2,11]. Therefore, improving the environmental design of urban public spaces to alleviate psychological stress in the elderly under severe and fluctuating climate conditions, enhancing their psychological resilience, and improving their quality of life has become an urgent issue to address [12,13,14].

Research indicates that elderly individuals are more susceptible to mental health problems [15], which are often accompanied by other diseases or life events associated with aging [14]. Furthermore, the physical functions of the elderly have undergone significant changes, affecting the relationship between sensory environments and psychological recovery. For instance, the efficiency of defense mechanisms against cold and heat has declined [16], and hearing loss is more common [17].

As an integral part of the daily lives of the elderly, urban public spaces can effectively reduce their psychological burden by creating restorative experiences [18,19]. These spaces not only provide natural landscapes and opportunities for social interaction but also significantly promote psychological recovery in the elderly through suitable environmental conditions [20,21]. Studies have shown that sensory environments, through stimuli such as visual, auditory, tactile, olfactory, and gustatory experiences, directly influence the emotions, psychological states, and behaviors of the elderly, having a profound impact on their quality of life [22,23,24]. Therefore, exploring how to enhance the psychological resilience of the elderly through the optimization of multisensory environmental design in urban public spaces holds significant theoretical and practical importance.

In regions with severe cold climates, these environmental challenges are even more complex. The extreme low temperatures and snowfall in winter limit the frequency of outdoor activities, increasing feelings of social isolation, while the cold environment may exacerbate physiological stress in the elderly population [25,26]. At the same time, the high temperatures and humidity in summer may induce heat stress, exacerbating psychological and physical discomfort [27,28]. Consequently, the elderly population faces unique health challenges under these extreme climate conditions.

1.2. Literature Review

1.2.1. The Psychological Restorative Potential of Urban Public Spaces

Urban public spaces possess significant potential for psychological restoration, especially among older adults. Kaplan’s restorative environment theory suggests that urban public spaces containing natural elements, such as green spaces, open views, and natural light, enhance psychological recovery capabilities [29]. Ulrich et al. further demonstrated that exposure to natural environments significantly reduces psychological stress, a benefit that is particularly effective for the elderly [30]. Moreover, Tzoulas et al. found that the use of urban green spaces not only improved residents’ mental health but also promoted overall social welfare [31].

For older adults, the psychological restorative effects of green spaces are especially prominent. Research by Sugiyama and Thompson indicates that green spaces within communities significantly increase outdoor activity frequency among the elderly, thereby improving their mental health [32]. Hunter et al. noted a significant positive correlation between the frequency of elderly use of urban green spaces and their well-being, particularly in high-stress urban environments [33]. Additionally, Kweon et al. discovered that older adults living in green space-rich communities exhibited higher levels of mental health compared to those in areas with fewer green spaces [34].

Environmental quality is also a key factor in urban public spaces’ impact on psychological recovery. Jiang et al. found that the higher the Green View Index in a city, the more pronounced the psychological restorative effects, especially among older adults [20]. Gidlöf-Gunnarsson and Öhrström’s research showed that controlling environmental noise and enhancing visual aesthetics not only alleviates psychological stress in older adults but also enhances their sense of well-being [35].

1.2.2. The Impact of Cold Climates on Urban Public Spaces

Cold climates significantly limit the frequency of use and psychological restorative functions of urban public spaces. Research by Nikolopoulou and Steemers revealed that extreme cold temperatures and snow greatly reduce the use of public spaces, particularly affecting older adults [36]. Chen et al. further pointed out that temperature, wind speed, and humidity in cold climates are key factors influencing the use of public spaces by older adults. Cold weather often leads to a reduction in outdoor activity, exacerbating mental health issues among the elderly [37]. Moreover, the restorative effects of snow-covered vegetation during winter are diminished compared to the greenery of summer [38].

Cold climates also have a direct negative impact on the mental health of older adults. Su et al. found that cold weather significantly increases psychological stress levels in the elderly, exacerbating feelings of loneliness and anxiety [39]. Cojocaru et al. showed that cold weather not only reduces outdoor activity time but also worsens the mental state of older adults, resulting in increased symptoms of depression and anxiety [40]. These findings highlight the heightened vulnerability of older adults’ mental health in cold climates, underscoring the need for public space designs that cater to the specific needs of this demographic.

To address the challenges posed by cold climates in urban public spaces, researchers have proposed various adaptive design strategies. Knez and Thorsson found that improving microclimate design, such as adding windbreaks and shading structures, can significantly enhance the comfort and usability of public spaces [27]. Hwang et al. emphasized that designing appropriate thermal environments, especially by providing shelters and heating facilities, helps maintain high usage rates of public spaces during cold seasons and reduces psychological stress responses in older adults [41].

1.2.3. The Combined Effects of Multi-Sensory Environments on Psychological Restoration

The integrated design of multi-sensory environments plays a significant role in psychological restoration. Elsadek et al. noted that an appropriate thermal environment can significantly reduce physiological stress, creating favorable conditions for the positive effects of other sensory stimuli, thereby enhancing psychological recovery [42]. Song et al.’s research revealed that the combination of visual and olfactory stimuli may produce additive effects, further enhancing psychological restoration [43]. Meanwhile, the combination of visual and auditory stimuli can have either enhancing or interfering effects, depending on the specific context [44]. Additionally, Park et al. found that presenting visual imagery could enhance the restorative effects of water sounds on psychological recovery [45]. However, certain sensory stimuli combinations may produce different outcomes. For example, Qi et al. found that in the presence of birdsong, visual and olfactory stimuli did not significantly contribute to psychological restoration [46]. Nevertheless, Herzog and Strevey’s experimental research demonstrated that the combination of soundscapes and visual landscapes could significantly enhance psychological recovery, especially among older adults, where the harmonious combination of visual and auditory stimuli effectively reduced mental fatigue [47].

Research suggests that the integration of multi-sensory environments can maximize psychological restorative effects. Ba et al. discovered that combining olfactory stimuli, such as floral scents, with visual landscapes significantly enhances relaxation [48]. Bai and Jin’s study, which considered urban environmental factors, confirmed the positive role of multi-sensory environments in enhancing psychological restoration among older adults in cold climates [49]. These findings further support the importance of multi-sensory design in cold climates and emphasize the necessity of promoting such designs among older adults. Joynt et al.’s research indicated that although deciduous vegetation barriers may be visually appealing, their effectiveness in noise reduction is relatively limited [50]. Ebenberger et al. showed that during hot summer months, the thermal environment’s importance for psychological restoration even surpasses that of aesthetically pleasing forest stands [51].

1.3. Aim of Research

Existing studies have extensively explored the impact of sensory environments on psychological restoration, but they have primarily focused on younger or middle-aged populations in temperate climates, with insufficient attention paid to the elderly in cold regions. The seasonal variations in cold regions have a particularly significant effect on sensory environments, as snow cover and low temperatures in winter may alter the restorative effects of visual and thermal environments. Therefore, the core objective of this research is to investigate how thermal environments, in conjunction with other sensory factors, influence the psychological restoration of older adults under cold conditions.

While considerable progress has been made in understanding these dynamics within temperate climates, a significant research gap remains regarding how sensory environments in cold regions collectively impact the psychological restoration of the elderly across different seasons. In particular, there is a lack of in-depth analysis on how seasonal factors alter the effects of sensory stimuli on the psychological recovery of older adults. This research aims to address this gap by focusing on the influence of sensory environments on the psychological restoration of elderly individuals in urban public spaces across different seasons in cold regions.

We hypothesize that, under specific environmental conditions in cold regions, sensory stimuli across different seasons will have a significant impact on the psychological restoration of the elderly. By thoroughly analyzing visual, auditory, and thermal environments, this research will explore how these sensory factors interact across seasons to affect the psychological restoration of the elderly population. This will not only deepen our understanding of the relationship between sensory environments and psychological restoration under cold climate conditions, but also provide practical guidance for optimizing the design of urban public spaces to maximize psychological health benefits for older adults.

Figure 1 illustrates the conceptual framework of this research, aiming to identify and analyze the roles of these key environmental factors across different seasons. This framework is intended to provide scientific guidance for the design of urban public spaces in cold regions, effectively improving the mental health of the elderly population.

2. Materials and Methods

2.1. Study Sites

Harbin is located in northeastern China (125°42′–130°10′ E, 44°04′–46°40′ N, elevation: 132–200 m) and experiences a cold and snowy winter climate, classified as a severe cold region [52]. According to the seasonal climate classification by the China Meteorological Administration, winter in Harbin begins in October and lasts until early April of the following year [53]. Based on data from the National Meteorological Information Center of China (1981–2010, Figure 2), the average monthly temperature in Harbin during winter is −10.4 °C, with an average relative humidity of 64.8%, and the extreme minimum temperature can reach as low as −37.7 °C [54]. As of the end of 2020, the proportion of the population aged 60 and above in Harbin had reached 24.1%, indicating a significant and increasingly severe aging trend [55]. Therefore, Harbin serves as an important case study for exploring the impact of urban public space environments on the psychological restoration of the elderly.

This study selects urban public spaces in Harbin with distinct characteristics in terms of visual, thermal, and acoustic environments for field experiments. Due to research constraints, the study focuses primarily on urban public spaces, excluding commercial and industrial areas. Although the coverage of urban public spaces is limited, most daily activities of the residents are concentrated in these areas, making the study of their impact on residents’ well-being and psychological restoration highly valuable. Little is known about the potential of different types of urban public spaces (such as parks, streets, and residential community plazas) to promote health and well-being across different seasons. Therefore, three common types of urban public spaces in Harbin were selected as survey sites.

To avoid residual effects, participants visited one of six different environmental settings located in Harbin on the same day. Specifically, these six study sites represent different types of urban public spaces (Figure 3). A1 and A2 are situated in two parks: A1 is a relatively enclosed area surrounded by numerous trees, shrubs, and lawns, with a moderate number of visitors and proximity to a major traffic road; A2 is a relatively open area with a variety of trees, shrubs, and grass, where the number of people is relatively high, but it is located far from major traffic roads. B1 and B2 are located on two streets: B1 is a relatively open area with some buildings, a few large trees, and naturally growing grass, with a moderate number of people and frequent vehicle traffic; B2 is a relatively enclosed area surrounded by many trees, with a higher density of people and situated adjacent to a major traffic road. C1 and C2 are located in two residential community plazas: C1 is a relatively open area with some buildings, trees, shrubs, and a small lawn, adjacent to a secondary traffic road; C2 is a relatively enclosed area with numerous buildings, trees, shrubs, and lawns, adjacent to a major traffic road.

2.2. Field Survey

To comprehensively assess the psychological recovery of the elderly across different seasons, we conducted a field survey spanning five sessions over four seasons. These surveys took place on 16 May 2021 (spring), 6 July 2021 (summer), 24 October 2021 (autumn), 21 December 2020 (winter), and 1 December 2021 (winter). The survey team consisted of 12 professionally trained researchers, with two researchers assigned to each location to ensure consistency and accuracy in data collection.

The survey aimed to collect both demographic information and data on psychological recovery using the Recovery Outcome Scale (ROS). The questionnaire was divided into two parts. The first part gathered personal information, including gender and age. The second part focused on assessing psychological recovery using the ROS scale, which consists of six items. This scale, validated in prior studies [56,57], measures perceived recovery outcomes and is considered reliable and effective [58]. The six items of the ROS scale include three related to relaxation and calmness (“I feel restored and relaxed”, “I feel calm”, “I have enthusiasm and energy for my everyday routines”), one reflecting attentional recovery (“I feel focused and alert”), and two associated with mental clarity (“I can forget everyday worries”, “My thoughts are clear”) [59]. Respondents answered these items on a five-point Likert scale ranging from −2 (very little) to 2 (a lot). The average score across the six items was used to quantify the perceived recovery effects in the elderly population.

Respondents were randomly selected and surveyed under relatively stable conditions. All participants had normal vision and hearing and were capable of completing the questionnaire independently. To protect personal privacy and ensure the authenticity of the survey results, all responses were anonymous, and respondents were required to answer all questions. Each questionnaire was completed within 10 min.

A total of 1093 respondents participated in this study, with 229 in spring, 380 in winter, 240 in autumn, and 244 in summer. The age range of respondents was between 60 and 94 years. The largest proportion of respondents fell within the 60–69 age group, accounting for approximately 50% across all seasons. In contrast, the 80 and above age group was the smallest, ranging between 10.48% and 17.08%. The proportion of male respondents slightly exceeded that of female respondents, with a difference ranging from 4.10% to 17.90%.

The reliability of the ROS scale was evaluated using Cronbach’s Alpha, which ranged from 0.938 to 0.968 across the different seasons. According to Landis and Koch’s [60] criteria, a Cronbach’s Alpha above 0.801 indicates good internal reliability, confirming that the ROS scale used in this study was highly reliable.

2.3. Measures

Environmental measurement was divided into thermal environment testing and audiovisual environment collection (360-degree panoramic photos and videos). Measurement data can be found in Appendix A 

Table A1. Characteristics of the spring environment.

	A1	A2	B1	B2	C1	C2
Tg (°C)	25.164	28.133	28.767	25.853	34.349	24.178
Ta (°C)	21.103	22.606	23.284	22.048	23.533	22.002
v (m/s)	0.536	0.795	0.997	0.997	0.615	0.525
RH (%)	33.222	28.401	29.652	27.969	30.848	29.327
Tmrt (°C)	33.053	40.029	42.834	35.728	50.05	28.301
UTCI (°C)	23.622	26.106	26.992	24.095	29.716	22.585
Panoramic Green View Index (%)	24.388	11.525	6.615	24.340	8.353	6.590
Average Number of People	6.211	40.526	6.684	7.158	1.167	0.421
Average Number of Vehicles	0.842	0.000	28.053	11.263	4.000	0.000
Proportion of Snow (%)	4.4	2.1	6.2	7.6	2.4	12.4
LA10–LA90	11.084	8.048	11.263	11.369	16.642	19.252
dB(C)–dB(A)	4.123	5.216	5.393	9.779	14.618	13.674
N10–N90	5.672	3.199	6.417	8.778	4.185	5.521
Proportion of Traffic Noise (%)	56.7	0.0	93.3	73.3	40.0	3.3
Proportion of Conversation Noise (%)	40.0	96.7	60.0	43.3	20.0	26.7

,

Table A2. Characteristics of the winter environment.

	A1	A2	B1	B2	C1	C2
Tg (°C)	−8.390	−7.309	−8.177	−8.222	−9.204	−8.659
Ta (°C)	−10.265	−9.364	−10.015	−9.994	−10.460	−10.648
v (m/s)	0.285	1.129	1.451	0.957	1.317	1.034
RH (%)	58.783	55.173	58.022	58.97	60.202	60.508
Tmrt (°C)	−4.985	1.547	0.896	−1.181	−3.206	−0.484
UTCI (°C)	−8.688	−9.949	−12.504	−10.269	−13.400	−11.135
Panoramic Green View Index (%)	24.388	11.525	6.615	24.340	8.353	6.590
Average Number of People	6.211	40.526	6.684	7.158	1.167	0.421
Average Number of Vehicles	0.842	0.000	28.053	11.263	4.000	0.000
Proportion of Snow (%)	4.4	2.1	6.2	7.6	2.4	12.4
LA10–LA90	11.084	8.048	11.263	11.369	16.642	19.252
dB(C)–dB(A)	4.123	5.216	5.393	9.779	14.618	13.674
N10–N90	5.672	3.199	6.417	8.778	4.185	5.521
Proportion of Traffic Noise (%)	56.7	0.0	93.3	73.3	40.0	3.3
Proportion of Conversation Noise (%)	40.0	96.7	60.0	43.3	20.0	26.7

,

Table A3. Characteristics of the autumn environment.

	A1	A2	B1	B2	C1	C2
Tg (°C)	13.486	12.251	19.146	15.769	12.952	12.764
Ta (°C)	12.326	12.690	14.037	14.086	13.085	12.838
v (m/s)	0.176	0.475	1.107	0.228	0.509	0.709
RH (%)	40.100	39.949	37.550	36.562	39.608	39.071
Tmrt (°C)	15.062	11.780	33.874	18.233	13.002	13.064
UTCI (°C)	13.354	11.916	18.689	15.213	12.579	11.766
Panoramic Green View Index (%)	24.388	11.525	6.615	24.340	8.353	6.590
Average Number of People	6.211	40.526	6.684	7.158	1.167	0.421
Average Number of Vehicles	0.842	0.000	28.053	11.263	4.000	0.000
Proportion of Snow (%)	4.4	2.1	6.2	7.6	2.4	12.4
LA10–LA90	11.084	8.048	11.263	11.369	16.64	19.252
dB(C)–dB(A)	4.123	5.216	5.393	9.779	14.618	13.674
N10–N90	5.672	3.199	6.417	8.778	4.185	5.521
Proportion of Traffic Noise (%)	56.7	0.0	93.3	73.3	40.0	3.3
Proportion of Conversation Noise (%)	40.0	96.7	60.0	43.3	20.0	26.7

and

Table A4. Characteristics of the summer environment.

	A1	A2	B1	B2	C1	C2
Tg (°C)	29.276	32.257	31.560	29.480	36.805	32.021
Ta (°C)	28.377	29.196	29.278	29.323	30.310	28.905
v (m/s)	0.178	0.252	1.413	1.001	0.491	0.665
RH (%)	68.084	63.074	60.973	62.515	61.873	64.902
Tmrt (°C)	30.208	35.784	39.059	29.870	46.981	37.929
UTCI (°C)	30.576	32.312	31.945	30.276	35.900	32.484
Panoramic Green View Index (%)	47.568	27.458	18.510	40.931	24.795	33.487
Average Number of People	9.632	48.842	5.579	17.737	13.211	2.842
Average Number of Vehicles	0.000	0.000	33.526	9.053	2.000	0.000
Proportion of Snow (%)	12.084	6.628	9.801	5.602	11.455	7.558
LA10–LA90	6.111	9.036	7.248	8.524	9.847	13.474
dB(C)–dB(A)	6.583	4.858	5.642	5.171	2.076	2.983
N10–N90	26.7	26.7	90.0	93.3	0.0	10.0
Proportion of Traffic Noise (%)	40.0	70.0	56.7	76.7	93.3	36.7
Proportion of Conversation Noise (%)	47.568	27.458	18.510	40.931	24.795	33.487

. The thermal environment testing was conducted simultaneously with the questionnaire survey, from 8:00 a.m. to 5:00 p.m. However, the audiovisual environment collection took place the day after the survey, from 10:00 a.m. to 2:00 p.m. This schedule was designed to prevent any interference from participant gatherings or conversations during the questionnaire survey that could affect the audiovisual data collection.

A 5.7K high-definition spherical panoramic camera (Insta 360 ONE X2, Insta 360, Shenzhen, China) with a time resolution of 30 fps and a bitrate of 95 Mbps was used to capture high-quality images within a 360° field of view. The device is equipped with four built-in microphones capable of capturing 360-degree panoramic sound. In accordance with recommendations (ISO 12913-2 [61]), the audiovisual data were collected for 5 min, and the camera tripod height was adjusted to 1.5 m to meet the standard. In the areas where the thermal environment testing equipment was installed (see

Table 1. Characteristics of thermal environment measurement instruments.

Thermal Environment Parameter	Instrument Name	Range	Accuracy	Sampling Rate
Wind speed	Kestrel 5500 weather station, Kestrel, Boothwyn, PA, USA	0.4~40 m/s	±0.1 m/s	1 min
Black globe temperature	BES-01 temperature recorder, Harbin, China	−30~50 °C	±0.5 °C	1 min
Air temperature	BES-02 temperature and humidity recorder, Harbin, China	−30~50 °C	±0.5 °C	1 min

), there were no other interfering crowds present.

2.3.1. Visual Environment Stimulus Variables

We selected two types of visual environment stimulus variables, including natural feature variables and human-made variables. Based on previous research [62,63,64], we selected the panoramic green view index (PGVI) and the proportion of snow (PS) as the natural feature variables. The PGVI is calculated by the proportion of green pixels in panoramic images [65,66]. In this study, we used Adobe Photoshop 2020 software to obtain data on the green pixels and the total pixels, and the PGVI was calculated using the following formula:

$$\mathrm{P}\mathrm{G}\mathrm{V}\mathrm{I}=\left({\displaystyle \frac{{Area}_{{green}_{panorama}}}{{Area}_{{total}_{panorama}}}}\right)\times 100\%.$$

(1)

The calculation of the snow coverage (during winter) is similar to that for the panoramic green view index, measuring snow coverage by the number of pixels. Concerning the research conditions in the urban public spaces, we chose to investigate the average number of people and the average number of vehicles as the human variables. The average number of people is used to describe the population density in the field of view, while the average number of vehicles describes the vehicle density in the field of view. To conduct this investigation, we trained two researchers to observe the number of people and vehicles in 360-degree panoramic videos. We divided a 5 min, 360-degree video into 30 segments of 10 s each and then calculated the average number of people and the average number of vehicles during the 5 min period.

2.3.2. Thermal Environment

Based on previous research experience, this study selected on-site test parameters and calculated the parameters as stimulus variables for the thermal environment [67]. The on-site measurement parameters included the air temperature (°C), relative humidity (%), black globe temperature (°C), and wind speed (m/s). Table 1 lists the characteristics of the testing instruments used. The diameter of the black globe is 0.08 m, and the surface material has a scattering coefficient of 0.95. Before these instruments were used, they were all calibrated. To avoid interference from solar radiation and wind, the temperature and humidity recorders were placed inside a radiation-resistant aluminum shield. We mounted the instruments on a tripod at a height of 1.5 m above the ground. The selection of these instruments complied with ASHRAE 55-2017 [68] and ISO 7726 [69]. Table 1 provides detailed information about each instrument, including their range, accuracy, and measured parameters.

In this study, the selected calculated parameters included the mean radiant temperature (T_mrt) and the universal thermal climate index (UTCI). The mean radiant temperature (T_mrt) is a key indicator for assessing urban radiation and thermal environments [70]. In this work, the T_mrt is calculated using the following formula:

$$T_{mrt}={[{\left(T_g+273.15\right)}^4+{\displaystyle \frac{1.10{\times 10}^8\times v^{0.6}}{\epsilon \times { D}^{0.4}}}(T_g-T_a\left)\right]}^{1/4}-273.15,$$

(2)

where T_g represents the black globe temperature, T_a represents the air temperature, ε represents the emissivity of the black globe surface, D represents the diameter of the black globe, and v represents the air velocity.

The universal thermal climate index (UTCI) is an indicator used to assess the thermal environment. It comprehensively considers factors such as the temperature, humidity, and wind speed to provide an integrated thermal index. The UTCI has been widely adopted in outdoor thermal environment assessments [67], reflecting the physiological responses of humans to multidimensionally defined actual thermal conditions [71]. In this study, we used BioKlima 2.6 software to calculate the UTCI.

2.3.3. Acoustic Environment

Based on previous studies [72,73], we analyzed the stimulus variables of the acoustic environment, including psychoacoustic indices and sound proportions. First, to check the background noise at the test locations, we used the audio files recorded from 360-degree panoramic videos (5 min) to calculate the psychoacoustic parameters, including L_A10–L_A90, dB(C)–dB(A), and N₁₀–N₉₀. The acoustic index L_A10–L_A90, associated with the acoustic climate, has been widely used to detect noise fluctuation characteristics [74]; dB(C)–dB(A) considers both C-weighted and A-weighted filters, indicating the amount of low-frequency energy in the noise [75]; N₁₀–N₉₀ represents the difference between N₁₀ and N₉₀ [76], which are the excess percentages of the psychoacoustic parameter loudness (N) calculated according to ISO 532-1 [77]. All the psychoacoustic parameters were calculated using commercial software (BK Connect 2018, Brüel & Kjær, Nærum, Denmark).

Secondly, to calculate the proportion of traffic noise and conversation noise, we asked two professional students with normal hearing to listen to the 5 min audio. They observed in 10-s intervals and recorded the presence of traffic noise and conversation noise within each 10-s segment. By statistically analyzing these records, we were able to determine the proportion of traffic noise and conversation noise in the entire audio.

2.4. Data Analysis

This study used IBM SPSS Statistics v26 for data aggregation and analysis. The analytical process consisted of several steps. First, a one-way analysis of variance (ANOVA) was conducted to examine whether there were significant differences in psychological restoration among the elderly across different seasons and study locations. The effect size of environmental factors on psychological restoration was also assessed. Next, Pearson correlation analysis was performed to explore the relationships between environmental stimulus variables in urban public spaces (including visual, thermal, and acoustic environments) and psychological restoration (ROS scores).

Based on the results of the above analyses, a psychological restoration assessment model was further constructed. Stepwise multiple linear regression was employed to quantify the explained variance (R²) of ROS scores attributed to visual, thermal, and acoustic stimuli across different seasons, thereby identifying the environmental factors that had a significant impact on ROS scores. To gain deeper insights into the importance of each variable within the model, the “relaimpo” package in R software (version 4.2.1) was used to calculate the relative importance (RI) of each variable. The total explained variance (R²) was decomposed using the LMG metric method, allowing for the contribution of each predictor variable to the psychological restoration effect to be determined, and identifying the most influential variables on psychological restoration.

Finally, a conditional probability method was applied to determine the likelihood of urban public spaces promoting or hindering psychological restoration under different ranges of environmental variable values. This helped identify which environmental conditions were more likely to influence psychological restoration, providing key insights for model interpretation and optimization.

3. Results

3.1. The Impact of Urban Public Spaces on Psychological Restoration in the Elderly

The impact of different seasons on the psychological restoration of the elderly exhibits significant differences, and these differences hold practical importance. The results of the one-way analysis of variance (ANOVA) presented in

Table 2. One-way ANOVA and effect sizes for the impact of different seasons on the psychological restoration of the elderly.

Season	F	p	Effect Size (η2)
Spring	25.036	<0.001 ***	0.360
Winter	18.765	<0.001 ***	0.201
Autumn	17.677	<0.001 ***	0.274
Summer	22.123	<0.001 ***	0.317

Note: * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; η² ≥ 0.04 = small effect; η² ≥ 0.25 = moderate effect; η² ≥ 0.64 = large effect.

provide detailed insights into this, including F-values, statistical significance (p-values), and effect sizes (η²). The study reveals that there are significant differences in the effectiveness of psychological restoration across different seasons. Spring has the highest effect size, reaching 0.360, indicating a significant moderate impact of the spring environment on the psychological restoration of the elderly. In contrast, the effect size for winter is 0.201, suggesting a smaller impact of the winter environment on psychological restoration. The effect sizes for autumn and summer are 0.274 and 0.317, respectively, indicating that urban public spaces during these seasons also have a moderate impact on the psychological restoration of the elderly. These results highlight that the psychological restoration effects of urban public spaces on the elderly vary by season, with certain seasons (such as spring) having a more pronounced impact.

3.2. Correlation between Psychological Restoration and the Environmental Stimulus Variables

To analyze the correlation between psychological restoration and the environmental stimulus variables (visual, thermal, and acoustic environments) across different seasons, Pearson correlation tests were used. According to the correlation analysis results shown in

Table 3. Correlation between psychological restoration and environmental stimulus variables.

Variable	Spring	Winter	Autumn	Summer
Visual Environment Index
Panoramic Green View Index	0.114	0.301 **	0.041	−0.081
Average Number of People	0.050	−0.102 *	0.298 **	0.458 **
Average Number of Vehicles	−0.480 **	−0.226 **	−0.393 **	−0.302 **
Proportion of Snow		−0.118 *
Thermal Environment Index
Black Globe Temperature	0.180 **	−0.017	−0.104	0.220 **
Air Temperature	−0.092	−0.057	0.002	0.051
Wind Speed	−0.350 **	−0.219 **	−0.062	−0.294 **
Relative Humidity	0.167*	0.076	−0.035	−0.009
Tmrt	0.095	−0.051	−0.150 *	0.148 *
UTCI	0.113	0.096	−0.106	0.196 **
Acoustic Environment Index
LA10−LA90	0.139 *	−0.028	−0.462 **	−0.014
dB(C)−dB(A)	0.055	−0.137 **	−0.472 **	−0.047
N10−N90	−0.104	−0.025	−0.115	−0.021
Proportion of Traffic Noise	−0.559 **	0.041	−0.218 **	−0.318 **
Proportion of Conversation Noise	0.148 *	−0.144 **	0.107	0.163 *

*, significance at the 0.05 level; **, significance at the 0.01 level.

, there is a significant relationship between the psychological restoration of the elderly (ROS scores) and the environmental stimulus variables, with seasonal differences. Firstly, throughout the year, the psychological restoration of the elderly is significantly negatively correlated with the average number of vehicles (p < 0.01, r < 0). Secondly, in winter, the psychological restoration is positively correlated with the panoramic green view index (p < 0.01, r = 0.301). In other seasons there is no correlation between the psychological restoration and the panoramic green view index. Additionally, summer has the most environmental stimulus variables positively correlated with psychological restoration, including the average number of people, the black globe temperature, the T_mrt, the UTCI, and the proportion of conversation noise. In winter, there are the most environmental stimulus variables negatively correlated with psychological restoration, including the average number of people, the average number of vehicles, the proportion of snow, the wind speed, the dB(C)–dB(A), and the proportion of conversation noise.

3.3. Psychological Restoration Assessment Model Based on the Visual, Thermal, and Acoustic Environments

In this study, we established psychological restoration assessment models for the elderly in different seasons to investigate the impact of various environmental stimulus variables on psychological restoration (ROS scores). Considering the potential autocorrelation between environmental stimulus variables, we used stepwise multiple linear regression, a method widely applied in previous research [78,79,80,81]. To detect multicollinearity among the independent variables in the model, we used the variance inflation factor (VIF) as the test indicator. According to previous studies, a VIF of <5 indicates no multicollinearity among the independent variables. The results (Table 3) show that the psychological restoration assessment model established in this study meets this condition.

Through the established psychological restoration assessment model (

Table 4. Psychological restoration assessment model in different seasons.

Season	Independent Variable	β (Standardized)	t	Sig.	VIF
Spring R2 = 0.398, adjusted R2 = 0.382	Proportion of Traffic Noise	−0.237	−2.477	0.014	3.386
	Wind Speed	−0.166	−2.744	0.007	1.356
	Black Globe Temperature	0.209	3.551	0.000	1.278
	Relative Humidity	0.133	2.247	0.026	1.295
	Average Number of Vehicles	−0.266	−2.865	0.005	3.179
	Proportion of Conversation Noise	0.133	2.327	0.021	1.198
Winter R2 = 0.200, adjusted R2 = 0.192	Average Number of Vehicles	−0.367	−7.288	0.000	1.189
	Average Number of People	−0.477	−7.561	0.000	1.866
	dB(C)−dB(A)	−0.339	−5.906	0.000	1.548
	Proportion of Snow	−0.187	−3.276	0.001	1.536
Autumn R2 = 0.273, adjusted R2 = 0.267	dB(C)−dB(A)	−0.375	−6.222	0.000	1.185
	Average Number of Vehicles	−0.245	−4.069	0.000	1.185
Summer R2 = 0.289, adjusted R2 = 0.284	Average Number of People	0.436	7.998	0.000	1.006
	Proportion of Traffic Noise	−0.283	−5.203	0.000	1.006

), we found that the variance explanation rate was highest in spring (R² = 0.398, adjusted R² = 0.382) and lowest in winter (R² = 0.200, adjusted R² = 0.192). This means that the visual–thermal–acoustic environment has a higher impact on the psychological restoration of the elderly in spring, while the impact is lower in winter. Furthermore, the psychological restoration assessment model includes the most environmental variables in spring (proportion of traffic noise, wind speed, black globe temperature, relative humidity, average number of vehicles, and proportion of conversation noise). Thus, when studying psychological restoration in spring, the largest number of stimulus variables needs to be considered. However, in autumn, the model included the fewest environmental variables (dB(C)–dB(A) and average number of vehicles), indicating fewer stimulus variables to consider.

Environmental stimulus variables can either enhance or diminish the psychological restoration. On one hand, the effects of the environmental stimulus variables may be inconsistent across different seasons. For example, an increase in the average number of people in winter leads to diminished psychological restoration (Standardized Beta = −0.477), whereas an increase in the average number of people in the summer enhances psychological restoration (Standardized Beta = 0.436). On the other hand, the effects of environmental stimulus variables may be consistent. For instance, the proportion of traffic noise diminishes psychological restoration in both spring and summer (Standardized Beta = −0.237 and −0.283, respectively); dB(C)–dB(A) diminishes psychological restoration in both winter and autumn (Standardized Beta = −0.339 and −0.375, respectively). The average number of vehicles reduces psychological restoration in spring, winter, and autumn (Standardized Beta = −0.266, −0.367, and −0.245, respectively). Additionally, it is noteworthy that in spring, the black globe temperature, relative humidity, and the proportion of conversation noise all enhance psychological restoration (Standardized Beta = 0.209, 0.133, and 0.133, respectively).

3.4. Relative Importance of Environmental Stimulus Variables

To further analyze the relative importance of the environmental stimulus variables on psychological restoration across different seasons, the “relaimpo” package in R software (version 4.2.1) was used to rank the variance contribution rates. This involved decomposing the variance explanation rate of the psychological restoration models for different seasons into the variance contribution rates of each independent variable (Figure 4).

In spring, the assessment model included visual, thermal, and acoustic environmental stimulus variables. The impact of the visual and thermal environments on psychological restoration was almost the same, with R² contribution rates of 11.39% and 11.38%, respectively. The R² contribution rate of the visual environment came entirely from the average number of vehicles (ANV). The thermal environment had multiple influencing factors, including wind speed, black globe temperature, and relative humidity. The acoustic environment played a significant role in psychological restoration, with an R² contribution rate of 17.03%, of which the proportion of traffic noise was 15.14%, and the proportion of conversation noise was only 1.89%. Overall, in spring, the acoustic environment had a greater impact on psychological restoration compared to the visual and thermal environments.

In winter, autumn, and summer, the assessment models included visual and acoustic environment variables but did not include thermal environment variables. In winter and summer, the visual environment had a larger impact on psychological restoration (R² contribution rates of 15.77% and 19.91%, respectively), while the acoustic environment had a smaller impact (R² contribution rates of 4.27% and 9.04%, respectively). In autumn, the acoustic environment had a larger impact (R² contribution rate of 17.07%), and the visual environment had a smaller impact (R² contribution rate of 10.28%).

3.5. The Impact of Environmental Variable Ranges on Psychological Restoration

To better understand how environmental stimulus variables influence the psychological restoration of the elderly in urban public spaces, a conditional probability method was applied to analyze the effects of different variable ranges. Conditional probability is a statistical method that calculates the probability of one event occurring given that another event has already occurred, and it has been widely used in similar research [82,83]. Specifically, conditional probability helps us evaluate how the probability of psychological restoration is affected by the range of environmental variable values under specific conditions. For instance, if the goal is to study event A (psychological restoration), and event B (the environmental variable value) has already occurred, the probability of A occurring given B is represented as P(A|B).

In this study, we defined spaces with ROS values in the range of [−2, 0] as environments that “hinder” psychological restoration, while spaces with ROS values in the range of (0, 2] were defined as environments that “promote” psychological restoration. For ease of analysis, environmental stimulus variables were divided into different ranges, such as wind speed in 0.5 m/s increments, percentage in 5% increments, and L_A10–L_A90 in 5 dB increments. Based on this, we further analyzed the probability of these variables promoting or hindering psychological restoration within different value ranges for the elderly.

According to the psychological restoration assessment criteria, urban public spaces with ROS values in the range of [−2, 0] were classified as environments that “hinder” psychological restoration, while spaces with ROS values in the range of (0, 2] were classified as environments that “promote” psychological restoration [49]. To more precisely analyze the impact of various environmental stimulus variables on psychological restoration, these variables were categorized into specific value ranges, such as wind speed in 0.5 m/s increments, percentage in 5% increments, and L_A10–L_A90 in 5 dB increments [84]. Based on these ranges, we analyzed the probability of these variables promoting or hindering psychological restoration and presented the conditional probability distributions for spring, summer, autumn, and winter separately (Figure 5, Figure 6, Figure 7 and Figure 8). These figures clearly compare the conditional probabilities of different environmental variables promoting versus hindering psychological restoration.

In spring (Figure 5), environmental factors that hinder psychological restoration are primarily concentrated in the medium to high range. For example, when wind velocity (v) is within the ranges of (0.5 m/s, 1 m/s] and (1 m/s, 1.5 m/s], the probability of hindering psychological restoration is 44.32% and 25.00%, respectively. Similarly, when the black globe temperature (T_g) is in the ranges of (10 °C, 15 °C] and (15 °C, 20 °C], the probability of hindering psychological restoration is 30.77% and 26.92%, respectively. These results indicate that higher wind velocity and T_g may cause discomfort or stress among older adults, thereby inhibiting their psychological restoration. Additionally, when relative humidity (RH) is within the ranges of (15%, 20%] and (20%, 25%], the probability of hindering psychological restoration is 46.15% and 38.46%, respectively. In contrast, lower wind velocity ([0 m/s, 0.5 m/s] and (0.5 m/s, 1 m/s]) and lower proportion of traffic noise (PTN) ([0%, 5%]) are more conducive to promoting psychological restoration, with promotion probabilities of 37.35%, 51.81%, and 53.01%, respectively. Notably, when the average number of vehicles (ANV) is in the range of [0, 5], the probability of promoting psychological restoration reaches 97.59%. These findings suggest that reducing traffic noise and controlling environmental wind velocity may have a positive impact on the mental health of older adults.

In winter (Figure 6), the data show that environmental factors hindering psychological restoration are primarily concentrated in higher proportions of snow (PS) and larger dB(C)–dB(A) ranges. Under these conditions, the probability of urban public spaces hindering psychological restoration significantly increases, especially when the PS reaches the range of (5%, 10%], where the probability is as high as 39.79%. This phenomenon may be due to the presence of snow and the larger noise differences inducing discomfort and stress, thereby inhibiting psychological restoration. Conversely, conditions that promote psychological restoration include combinations of a low average number of vehicles (ANV) ([0, 5] range) and low dB(C)–dB(A) (within the [0 dB, 5 dB] range). Under these conditions, the probability of psychological restoration in urban public spaces significantly increases, reaching 78.46% and 63.08%, respectively. This suggests that reducing the number of vehicles and controlling the dB(C)–dB(A) are critical for enhancing psychological restoration in urban public spaces during the winter season.

In autumn (Figure 7), the data indicate that when the dB(C)–dB(A) range is within (5 dB, 10 dB] and (10 dB, 15 dB], the probability of urban public spaces hindering psychological restoration is 79.17% and 12.50%, respectively. Simultaneously, when the average number of vehicles (ANV) is in the lower range of [0, 5], the probability of hindering psychological restoration is 52.08%; however, when the ANV is within the (25, 30] range, the hindrance probability is 47.92%. In conditions conducive to psychological restoration, when the dB(C)–dB(A) is in the [0 dB, 5 dB] range, the probability of promoting psychological restoration is 73.33%, and when within the (5 dB, 10 dB] range, the promotion probability is 20.00%. Notably, when the ANV is in the [0, 5] range, the probability of urban public spaces promoting psychological restoration reaches 100%. This demonstrates that a very low average number of vehicles has a significantly positive impact on psychological restoration. This analysis indicates that in autumn, dB(C)–dB(A) and the number of vehicles are key environmental factors affecting psychological restoration in urban public spaces. Controlling dB(C)–dB(A) and reducing vehicle numbers, especially maintaining extremely low traffic flow, can significantly enhance psychological restoration in urban public spaces.

In summer (Figure 8), the data show that environmental factors hindering psychological restoration are mainly concentrated in the medium to high value ranges. For instance, when the average number of people (ANP) is within the ranges of (5, 10] and (10, 15], the probability of hindering psychological restoration is 39.39% and 3.03%, respectively. Additionally, when the proportion of traffic noise (PTN) is within the range of (90%, 95%], the probability of hindering psychological restoration is 25.76%, and within the range of (95%, 100%], it is 22.73%. In contrast, conditions that promote psychological restoration mainly occur at lower or higher extreme value ranges. For example, when PTN is within the range of [0%, 5%], the probability of promoting psychological restoration is 4.76%; whereas in the ranges of (35%, 40%] and (50%, 55%], the promotion probabilities reach 90.48% and 73.81%, respectively.

These data suggest that lower dB(C)–dB(A), moderate proportions of traffic noise, and appropriate average numbers of people may facilitate psychological restoration in the elderly. In summary, in the summer environment, reasonable control of the average number of people and the proportion of traffic noise, especially within the medium-low or medium-high ranges, may positively influence the psychological restoration of elderly individuals in urban public spaces.

In conclusion, across different seasons, environmental variables that hinder psychological restoration in the elderly tend to be concentrated in the medium to high value ranges, such as higher wind velocity (v), dB(C)–dB(A), and average number of people (ANP). In contrast, environmental variables that promote psychological restoration are often found in lower or extreme value ranges, such as lower dB(C)–dB(A), lower average number of vehicles (ANV), and moderate average number of people (ANP). These findings suggest that in designing urban public spaces, priority should be given to controlling environmental variables within ranges that are conducive to psychological restoration. Furthermore, adjustments should be made according to seasonal variations to maximize psychological health and restoration outcomes for the elderly.

4. Discussion

4.1. The Seasonal Impact of Urban Public Spaces on Psychological Restoration

This study demonstrates that different environmental factors have significant effects on the psychological restoration of the elderly across the four seasons. First, the visual environment played a key role in all seasons, which aligns with Ulrich’s “Stress Recovery Theory”, which posits that natural landscapes effectively reduce individuals’ physiological stress responses [19]. In this study, greenery and open views were particularly beneficial for psychological restoration during spring and summer, consistent with Cao et al.’s research on the impact of urban green spaces on public physiological and psychological health under various weather conditions. Their findings suggest that the combination of greenery and open views significantly promotes psychological restoration, especially on sunny days [85]. Moreover, Kaplan’s Attention Restoration Theory (ART) highlights the importance of elements in natural environments for psychological restoration, especially in spring and summer, where vegetation enhances the restorative effect [18]. Li et al.’s study further supports this conclusion, showing that vegetation in visual environments significantly promotes emotional recovery, particularly in residential courtyard spaces where green vegetation has a pronounced effect on emotional health [86].

The study also found that the acoustic environment significantly influenced psychological restoration during spring and summer, particularly by reducing background noise such as traffic noise, which markedly improved the psychological well-being of the elderly. This finding is supported by Gidlöf-Gunnarsson and Öhrström’s research, which emphasizes the importance of noise control for psychological restoration [35]. Yang et al. further validated this conclusion, showing that optimizing the acoustic environment during spring and summer can enhance psychological restoration in public spaces [87].

The thermal environment in winter had a particularly pronounced effect on psychological restoration. The study revealed that in cold climates, an appropriate thermal environment not only improved the comfort of the elderly but also significantly enhanced their psychological state. This finding is consistent with Lam et al.’s research on the impact of outdoor thermal comfort on psychological well-being [88]. Additionally, Zhu et al.’s study indicated that designing landscapes with rich layers and extensive coverage can enhance restorative potential in cold climates, especially in winter [38]. This design strategy aligns with the findings of this study regarding the influence of the winter thermal environment.

4.2. The Psychological Restorative Effects of Environmental Stimuli

This study found that the effects of visual environments on psychological restoration varied significantly across seasons, particularly in spring and summer, where green spaces had a notably strong contribution to psychological restoration. This aligns with Kaplan’s concept of “soft fascination”, which suggests that mild stimuli in natural environments can help restore attention [89]. Additionally, the results of this study support the view of Kuo and Sullivan, who argued that exposure to natural environments helps alleviate mental fatigue and enhances cognitive function [90]. Specifically, for the elderly, observing green plants significantly enhanced their psychological restoration, further validating the broad applicability of this theory.

The influence of thermal environments on psychological restoration was also particularly evident. De Dear and Lamberts’ research indicates that thermal regulation in cold climates has a direct impact on psychological restoration [91]. This study found that providing a warm outdoor environment significantly improved the restorative experience for the elderly, which is consistent with Knez et al.’s findings, emphasizing the influence of cultural background and environmental attitudes on park thermal comfort and emotional responses, especially the role of warm climates in psychological restoration [92].

Moreover, this study highlighted the interactive effects of visual, thermal, and acoustic environments. The combination of visual and thermal environments significantly enhanced the restorative experience of the elderly. This finding is consistent with Bowler et al.’s discovery of the synergistic effects of multi-sensory environments [93]. Lindal and Hartig’s research also suggests that combining visual and soundscapes can further improve psychological restoration across different seasons [94].

4.3. Improvement Strategies and Urban Design Recommendations

This study provides several urban design strategies based on an empirical analysis of the effects of visual, thermal, and acoustic environments on the psychological restoration of the elderly in urban public spaces in cold regions. First, the design of urban public spaces should fully consider the impact of seasonal differences. The study found that the positive effects of the spring environment on psychological restoration were the most significant, while the winter environment was relatively weaker. Therefore, design adjustments should be made according to seasonal characteristics to maximize the psychological restoration benefits for the elderly. For example, in spring, increasing greenery and reducing traffic noise can enhance psychological restoration.

Secondly, the study suggests that different environmental factors should be prioritized according to the season. In spring, reducing traffic noise and controlling wind velocity should be the focus to improve psychological restoration; in summer, attention should be given to improving black globe temperature (T_g) and natural soundscapes; and in winter, the number of vehicles should be reduced, snow coverage (PS) controlled, and wind velocity managed to mitigate the negative effects on psychological restoration.

Lastly, considering the interactive effects of visual, thermal, and acoustic environments is crucial for enhancing the mental health of the elderly. The study shows that the acoustic environment contributes most to psychological restoration in spring, while the visual environment is more significant in winter and summer. Therefore, urban design should be optimized according to the seasonal importance of environmental factors to create a multisensory environment more conducive to psychological restoration.

In summary, urban public space design strategies in cold regions should focus on seasonal adjustments and the collaborative design of multisensory environments to better meet the psychological restoration needs of the elderly, thereby improving their quality of life.

5. Conclusions

This study empirically investigated the effects of visual, thermal, and acoustic environments on the psychological restoration of elderly individuals in urban public spaces in cold regions, focusing on Harbin. The main findings of the research are as follows:

Seasonal differences significantly impacted the psychological restoration of the elderly. Urban public spaces in spring had a moderate positive effect on psychological restoration (η² = 0.360), while the impact in winter was relatively weaker (η² = 0.201). These results suggest that urban space design should be adjusted according to seasonal characteristics to maximize the psychological restoration benefits for the elderly.

There was a significant association between environmental stimulus variables and psychological restoration. The data across all seasons showed a significant negative correlation between the average number of vehicles (ANV) and psychological restoration (p < 0.01, r < 0). In winter, the panoramic green view index (PGVI) was positively correlated with psychological restoration (p < 0.01, r = 0.301). Additionally, the black globe temperature (T_g) and proportion of conversation noise (PCN) in summer had a significant positive effect on psychological restoration, while the proportion of snow (PS) and wind velocity (v) in winter showed a negative effect, highlighting the diverse seasonal influences of environmental factors on psychological restoration.

The psychological restoration assessment model revealed different seasonal impacts on restoration outcomes. Stepwise multiple linear regression analysis found that environmental factors in spring had the strongest explanatory power for psychological restoration (R² = 0.398, adjusted R² = 0.382), while winter had the lowest (R² = 0.200, adjusted R² = 0.192). The main influencing factors in spring included the proportion of traffic noise (PTN), wind velocity (v), black globe temperature (T_g), and relative humidity (RH), while in winter, the average number of vehicles (ANV) and dB(C)–dB(A) had a more significant impact on psychological restoration.

The relative importance of environmental stimulus variables varied by season. In spring, the acoustic environment contributed the most to psychological restoration (17.03% of the R² contribution), while in winter and summer, the visual environment was more prominent (15.77% and 19.91%, respectively). This suggests that urban space design should prioritize environmental factors with the greatest restorative effects in each season.

Conditional probability analysis revealed the specific impact of certain environmental variables on psychological restoration. In spring, when wind velocity was within the range of (0.5 m/s, 1 m/s], the probability of urban public spaces hindering psychological restoration was 44.32%; however, when the proportion of traffic noise (PTN) was below 5%, the probability of psychological restoration increased to 53.01%. In winter, when the proportion of snow (PS) was between (5%, 10%], the probability of hindering psychological restoration was 39.79%; however, when the average number of vehicles (ANV) was below five, the probability of psychological restoration significantly increased to 78.46%. These findings indicate that environmental variables have different effects on psychological restoration within specific ranges during different seasons, and optimizing for these variables in design can maximize psychological restoration for the elderly.

In summary, this study deepened the understanding of the impact of urban public space environments on the psychological restoration of elderly individuals in cold regions through empirical analysis. The findings underscore the importance of adaptive urban space design based on seasonal variations to enhance the psychological health and restoration outcomes of the elderly. This study not only provides practical guidance for urban planners and designers but also offers scientific evidence for policymakers seeking to improve the quality of life for the elderly. Additionally, the study identifies several potential areas for future research. Future studies should further explore differences in individual environmental sensitivities and extend the scope of research to other age groups and geographic regions. In-depth research into the interaction between different environmental stimuli will also contribute to designing more effective urban public spaces to better meet the needs of the aging population and ensure maximum psychological restoration.