The Sustainable Evaluation and Improvement of Age-Friendly Outdoor Thermal Environments in Rural Xi’an: A Perspective on Spatiotemporal Variations in Elderly Daily Activity

Abstract

Elderly individuals in rural China are highly vulnerable to extreme weather events and temperature fluctuations due to inadequate infrastructure in the built environment and constrained economic conditions, thereby increasing their health risks. Outdoor spaces represent one of the primary daily activity settings for rural older adults. However, existing research rarely links spatiotemporal patterns of outdoor activities to evidence-based thermal environment optimization, leaving a critical knowledge gap for age-friendly and sustainable rural design. This study focuses on the spatiotemporal differentiation patterns of daily outdoor activities among elderly people aged 60 years and above in rural Xi’an, as well as the optimization of spatial variations in thermal environments. Using on-site interviews, thermal environment measurements, thermal comfort questionnaires, continuous thermal environment monitoring, and machine learning based on random forest, this study drew the following conclusions: (1) outdoor activities in winter were concentrated between 9:00–11:00 and 13:00–17:00, while in summer, they shifted to the morning and evening periods, namely 6:00–9:00 and 17:00–21:00. (2) Models for outdoor clothing adjustment, thermal sensation, and thermal acceptability among elderly residents were established. The calculated neutral temperature was 10.19 °C, with a 90% outdoor thermal acceptability range of 9.6–27.2 °C and an 80% outdoor thermal acceptability range of 6.2–30.6 °C. These findings differ from those documented in regions with distinct climate zones and geographical settings. This discrepancy stems from regional climatic features, lifestyle variations between urban and rural older adults, and differences in the thermal environment quality of elderly-oriented outdoor activity spaces. (3) In winter, the acceptable period of the Universal Thermal Climate Index (UTCI) at south-facing entrances (10:30–16:30) was significantly longer than that in the courtyard (13:30–14:00). In summer, the comfortable period in the courtyard (before 10:00 and after 20:00) was longer than that at north-facing entrances (before 09:00). A random forest model for thermal sensation was established, and the relative importance of each parameter influencing thermal sensation was analyzed. On this basis, priority improvement pathways and strategies for the thermal environment, as well as suggestions for the subjective adaptive behaviors of elderly residents, were proposed. The research results of this study can provide technical solutions for age-friendly thermal environment design in rural areas, thereby safeguarding the comfort, health, and social well-being of the elderly population in rural areas.

1. Introduction

With the continuous intensification of China’s population aging trend, ensuring the healthy and comfortable life of the elderly has become one of the important thematic issues of the era. Data from the seventh national population census of China show that the population aged 60 and above is 260 million, accounting for 18.9% of the total population [1]; it is estimated that by 2030, this proportion will reach around 25% [2]. The elderly differ from other populations in three aspects: physiological, psychological, and behavioral characteristics [3], which lead to their reduced ability to regulate responses to external temperature changes and make them more sensitive to temperature fluctuations [4]. In particular, elderly individuals with cardiovascular and cerebrovascular diseases are most vulnerable to the effects of high and low temperatures [5], resulting in higher incidence rates and mortality rates among the elderly compared to young people [6,7]. As noted in the World Report on Ageing and Health, the interactions every individual has with the environments they live in throughout their lifetime exert an influence on human functioning [8]; therefore, a thoughtfully crafted thermal environment that caters to the comfort and health requirements of older adults can boost their sense of well-being and reduce related health hazards [9].To better address the issue of age-friendly living environments, since 2017, departments such as the National Health and Family Planning Commission have successively issued the “Healthy Aging Plan,” which clearly proposes the goal of “promoting the construction of age-friendly environments” and aims to “create a safe, convenient, comfortable, and barrier-free age-friendly environment system” for the majority of the elderly [10]. Outdoor thermal comfort research injects core momentum into low-carbon ecological design by accurately mitigating the urban heat island effect and efficiently reducing energy consumption, creating safe, comfortable, and barrier-free outdoor spaces for the elderly and closely aligning sustainable urban construction with the livability needs of an aging society [11]. Basic research on sustainable building design should focus on the core needs of users, strive to create high-quality thermal environments that meet thermal comfort requirements, and further construct a human settlement ecological system that embodies people’s well-being and integrates livability with sustainability [9]. Age-friendly rural design must take into account a series of factors, including age-adaptability and comfort of the built environment, income constraints, construction practices, local governance capacity, and maintenance requirements. Exploring the thermal environment comfort of the elderly’s daily activity spaces is one of the fundamental tasks for the age-friendliness of sustainable human settlement environments.

Population aging in rural China is more severe than that in urban areas, with the proportion of the elderly population reaching as high as 23.81%—7.99 percentage points higher than that in urban areas [12]. Additionally, rural areas lag behind urban areas in several aspects, including planning/architectural design, building performance, and human settlement quality. In 2022, the 13th Five-Year Plan for Healthy Ageing proposed that it is necessary to promote the construction of age-friendly living environments and the age-friendly renovation of residential buildings for the elderly [13]. With the intensification of rural hollowing in China, an increasing number of the younger generation have migrated to urban areas, resulting in a growing population of left-behind elderly. This makes the elderly prone to psychological loneliness. Frequent daily interactions among elderly neighbors play a crucial role in alleviating the elderly’s loneliness and relieving their negative emotions [14,15]. Various types of outdoor spaces serve as important carriers for the elderly to engage in daily gathering and communicating, leisure and entertainment, as well as exercise and physical work. Their age-friendly level affects the elderly’s physical and mental health [3] and sense of well-being [16]. How to construct a human settlement environment that meets the rural elderly’s needs for social interaction, mental health, and environmental adaptation has become a key aspect of rural construction in the future [17]. Due to differences in economic conditions, living habits, the completeness of public activity facilities, transportation accessibility, and other aspects between urban and rural elderly, there are disparities in the spatiotemporal characteristics of daily activities [18], as well as thermal adaptation patterns and comfort needs [19,20], among rural elderly. Therefore, the improvement strategies for outdoor spaces for rural elderly’s daily activities cannot be directly referenced from urban schemes.

Since the elderly have diverse daily activity contents and occupy various types of spaces, accurately grasping their daily activity content and dynamic spatiotemporal trajectories is essential for comprehensively ensuring comfort during different activities. It also provides the basis for “time-segmented and zone-specific” precise evaluation and improvement of the thermal environment in outdoor daily activity spaces. Washburn [21] revised the Physical Activity Scale for community-dwelling elderly people, which classifies the elderly’s physical activities into leisure activities, household activities, and occupational physical activities. On this basis, some scholars have begun to study the activities among the elderly population, especially by analyzing the activities of individual elderly people from the perspectives of time and space [22]. Existing studies cover two core dimensions: on the one hand, they systematically examine the spatiotemporal patterns, intragroup heterogeneity, and driving determinants of daily activities among the elderly; on the other hand, focusing on the elderly’s demand for outdoor space utilization, they categorize the typical outdoor activity spaces for this group [23,24,25]. There are significant differences in activity venues and behavioral characteristics between urban and rural areas, especially in terms of activity time and the frequency of space use. Compared with cities, the types of outdoor activity spaces in rural areas are relatively limited—most elderly people take their own courtyards as rest spots or gather randomly for activities at certain locations in the village [18]. Given the significant differences in activity spaces and behavioral characteristics between urban and rural elderly groups, as well as the local characteristics of rural outdoor activity spaces, relevant scholars have carried out systematic research focusing on physical activity and public space planning for the rural elderly. They have classified the activity types of the rural elderly, analyzed the influencing factors of rural daily activity spaces, and conducted correlation analysis between the social attributes and activity characteristics of this group [26,27,28]. The specific research focuses and findings of each study are shown in

Table 1. Findings from the literature related to daily activities.

Authors	Location	Research Focus	Findings	Year	Reference
Wei	Urban Nanjing, China	Activity spatiotemporal characteristics and group differences.	The study analyzed the spatiotemporal characteristics and intragroup differences of the elderly’s shopping, leisure, and personal affairs activities, and explored the individual and environmental factors influencing such activities.	2012	[23]
Wu	Urban Ningbo, China	Temporal and spatial characteristics and seasonality of elderly activities.	The study found that the activity time of the elderly was relatively fixed and significantly affected by seasonal changes, and further revealed gender differences in their activity characteristics in winter.	2019	[24]
Fang	Urban Lhasa, China	Functional classification and regional characteristics of elderly outdoor activity spaces.	The study classified the typical outdoor activity spaces used by the elderly in Lhasa, analyzed the functions of religion, rest, entertainment, and care, and proposed optimization suggestions.	2019	[25]
Guan	Rural Chongqing, China	Influencing Factors influencing for the distribution of daily activity spaces.	The study found that villagers’ travel purposes, travel capabilities, and travel frequency respectively affected the distribution characteristics, spatial coverage, and demand differences of rural daily activity spaces in research targeting the physical activity characteristics of the rural elderly.	2021	[26]
Wu	Rural Ningbo (Nanyu New Village), China	Activity type classification for the rural elderly.	The study divided the physical activities among the rural elderly into four categories: health promotion, social interaction, daily living, and recreation and leisure.	2022	[27]
Leng	Rural Hailin (Xin’an Town), Yichun (Wudalianchi Town), Changchun (Qijia Town), Kaiyuan (Qingyunbao Town), Shenyang (Haoguantun Town), China	Correlation of social attributes and activity characteristics of the Elderly.	This study conducted correlation analysis between the social attributes and activity characteristics of the rural elderly in cold regions and put forward targeted public space planning strategies.	2015	[28]

.

To create a comfortable outdoor thermal environment, the outdoor thermal comfort needs of the elderly serve as the basis for the design and improvement of thermal environments. Ma [29] conducted a study on the thermal perception of elderly visitors in urban parks in Xi’an, analyzing the relationship between the elderly’s thermal perception and their age, as well as chronic diseases. Building on prior explorations in individual local study contexts of outdoor thermal comfort specific to distinct population groups, with a particular focus on the elderly, Fang [30], Wang [31], Pan [32], Peng [33], and Su [34] further targeted the differentiated characteristics of outdoor thermal environments and the thermal comfort demands of target populations across diverse climatic zones in China, including Xi’an, Dalian, Lanzhou, and Changsha.

In terms of outdoor thermal environment design and improvement, the outdoor open space environment is a complex integrated system composed of multiple factors, in which the thermal environment is affected by temperature, wind speed, and other factors [35]. Tian [36] investigated the factors influencing thermal perception of visitors. Zhou [37] and Wang [38] investigated the factors influencing the outdoor thermal comfort of the elderly in Shanghai and Guangzhou, discussed the outdoor thermal comfort demands of the elderly across different regions, and proposed optimization strategies for the outdoor thermal environment of activity spaces for the elderly. Shen [39] and Cherchi [40] took Xiyuan Xincun in Hefei, China, and Osida Village in Italy as their respective research objects, analyzed the current status of the thermal environment in these rural areas, and put forward targeted optimization recommendations for the outdoor thermal environment. Existing studies have put forward outdoor thermal environment optimization strategies mostly for urban areas, and there is a lack of research on such strategies for outdoor spaces in rural areas. Targeted and precise evaluations for different regions are required to ensure the elderly’s thermal comfort needs. The specific research focuses and findings of each study are shown in

Table 2. Findings from the literature related to thermal comfort.

Authors	Location	Research Focus	Findings	Year of Publication	Reference
Fang	Urban Guangzhou, China	Thermal comfort evaluation method.	The study evaluated the conditions and characteristics of outdoor thermal environments across different types of measurement point spaces in nursing homes in Guangzhou. It derived the neutral outdoor physiological equivalent temperature (PET) values and thermal comfort thresholds for the elderly.	2019	[30]
Wang	Urban Dalian, China	Analysis of thermal environment differences.	The study systematically quantified the differences in human thermal comfort under direct sunlight, tree shade, and building shade during the transition season in cold regions to provide a reference for the outdoor space design of universities in cold regions.	2022	[31]
Pan	Urban Lanzhou, China	Seasonal differences in outdoor thermal comfort among the elderly.	The study comparatively analyzed the thermal comfort evaluations among the elderly for three representative urban parks in Lanzhou City in summer and autumn, and clarified the seasonal characteristics of park thermal comfort.	2023	[32]
Peng	Urban Changsha, China	Outdoor thermal comfort status and thermal comfort evaluation.	The study investigated the outdoor thermal comfort conditions and corresponding thermal comfort evaluation of elderly care buildings in Changsha City.	2024	[33]
Su	Urban Xi’an, China	Thermal environment evaluation for specific activity types.	The study investigated the outdoor thermal comfort conditions among the elderly under different activity types.	2024	[34]
Tian	Urban Xi’an, China	Thermal perception influencing factors.	The study proposed thermal perception optimization strategies and corresponding recommendations for open scenic areas from the perspective of different population groups, based on visitors’ thermal benchmarks, meteorological characteristics, and the distribution of factors influencing thermal perception.	2022	[36]
Zhou	Urban Shanghai, China	Thermal environment optimization strategies.	The study investigated the effects of thermal environmental factors on the winter thermal comfort of the elderly in hot summer and cold winter regions and put forward corresponding optimization strategies.	2024	[37]
Wang	Urban Guangzhou, China	Quantitative analysis of influencing factors on thermal comfort.	The study investigated thermal comfort among the elderly and the current thermal environment in elderly care facilities in hot and humid regions, constructed an outdoor thermal comfort prediction model for the elderly, and proposed optimization strategies for the outdoor thermal environment of elderly activity areas.	2024	[38]
Shen	Rural Hefei (Xiyuan New Village), China	Thermal environment simulation and optimization.	The study used ENVI-met software to conduct a quantitative analysis of the current rural thermal environment, screened activity spaces suitable for the elderly based on healthy temperature thresholds, and put forward optimization suggestions.	2024	[39]
Cherchi	Rural Osida, Italy	Rural thermal environment optimization strategies.	The study analyzed the problem of outdoor thermal discomfort in summer in the rural area of Osida, identified solar radiation as the core influencing factor, and proposed measures to improve the thermal environment.	2025	[40]

.

To summarize, existing studies have examined issues related to the elderly from different perspectives. However, the spatiotemporal differentiation characteristics of the elderly’s daily activities have not been organically integrated with thermal comfort and thermal environment research, and the following limitations remain: first, studies on daily activity patterns mainly focus on urban areas, while the seasonal activity patterns and spatial use characteristics of the rural elderly remain unclear. Second, existing thermal comfort studies mostly focus on urban residential buildings, parks, and streets, with limited attention paid to the actual activity spaces used by the rural elderly. Third, current improvement strategies are relatively general and have not fully integrated the elderly’s daily activity patterns, adaptive thermal comfort needs, and key factors affecting thermal sensation. Although the thermal environment improvement strategies proposed in existing studies, including building envelope optimization, spatial planning and design, and plant configuration, possess certain universality, they lack age-friendly, precise assessment based on the temporal and spatial differentiation laws of the elderly’s activities and their adaptive thermal comfort needs. Xi’an is a typical representative city in China’s cold regions, featuring hot and rainy summers, as well as cold and dry winters. To effectively improve the thermal environment quality of outdoor activity spaces for rural elderly in this area, it is first necessary to accurately grasp the temporal and spatial differentiation laws of the elderly’s daily activities and their outdoor adaptive thermal comfort needs, followed by conducting precise “time-segmented and zone-specific” evaluations of the thermal environment levels of various outdoor spaces where the elderly engage in activities. Therefore, the objectives of this study are as follows:

• Identify the spaces for the elderly’s outdoor daily activities and the activity types conducted at different times in summer and winter, and explore the spatiotemporal differentiation patterns of the elderly’s activities.
• Identify the outdoor thermal response patterns of the elderly and establish comfort demand benchmarks for their outdoor activity spaces to align with the goals of climate-resilient construction and the sustainable development of urban–rural human settlements.
• Evaluate the thermal environment levels of the elderly’s outdoor activity spaces and propose effective improvement paths for uncomfortable conditions.

2. Materials and Methods

2.1. An Overview of the Technical Route

Figure 1 illustrates the overall research framework of this paper. First, a questionnaire survey was conducted on the daily activities among the elderly to clarify their activity time, activity types, and activity spaces. The spatiotemporal differentiation characteristics of activity contents were systematically analyzed, and their spatiotemporal distribution patterns were summarized, which defined a clear temporal and spatial scope for the subsequent research on thermal comfort and thermal environment.

Based on the preliminary survey results, this study carried out a thermal comfort survey in the spatial environments where the elderly most frequently conducted daily activities, so as to obtain the neutral temperature and thermal acceptability range of the elderly in Xi’an and establish a thermal environment evaluation scale. Meanwhile, continuous thermal environment monitoring was performed in typical daily activity spaces to collect basic thermal environment data on activity spaces used by the elderly in rural areas of Xi’an.

Using the data obtained from the thermal comfort survey, the main influencing factors of the outdoor thermal environment were explored, and the random forest algorithm was adopted to evaluate the priority of outdoor thermal environment optimization strategies. Finally, the thermal environment of the elderly’s activity spaces was comprehensively evaluated using the thermal environment evaluation scale. According to the priority of outdoor thermal environment influencing factors determined in this study, key influencing factors were accurately matched with corresponding optimization strategies, so as to achieve the targeted optimization and improvement of the thermal environment. In the discussion section, the results obtained from the above research steps are compared and analyzed with the findings from the relevant literature, and the rationality and innovation of the research conclusions are discussed in depth. Detailed descriptions of the specific research methods are provided in the subsequent chapters. Meanwhile, statistical analysis was performed using IBM SPSS Statistics 29.0 (IBM Corp., Armonk, NY, USA), and figures were plotted with OriginPro 2025 (OriginLab Corporation, Northampton, MA, USA) in this paper.

2.2. An Overview of the Study Site

Xi’an is the capital of Shaanxi Province, China, located in the central part of the Guanzhong Plain. It borders the Weihe River to the north, and is backed by the Qinling Mountains to the south, between 107°40′–109°49′ E and 33°42′–34°45′ N. According to the building climate zoning standard (GB 50176-2016) [41], Xi’an belongs to a cold region. According to the Köppen climate classification, it falls under the hot-summer, cold-winter humid continental climate zone (Dwa) [42]. It has a typical warm temperate semi-humid continental monsoon climate with four distinct seasons: winter is cold and dry, and summer is hot and rainy. The building climate zoning map is shown in Figure 2.

2.3. Investigation on Temporal and Spatial Patterns of Outdoor Daily Activities

Members of the research team conducted an extensive survey on the daily activity patterns of elderly people aged 60 and above in four villages—Songcun Village, Guocun Village, Nanqiangcun Village, and Luojiazhuang Village—in Chang’an District of Xi’an City, in July 2023 (summer) and January 2024 (winter) (Figure 3). The survey adopted a sampling method combining random field visits and voluntary participation. After confirming that the respondents were aged 60 and above and were willing to participate, the investigators conducted questionnaire interviews. Considering that rural elderly people may have declining vision and hearing, as well as relatively low educational levels, the questionnaires were completed by the investigators through one-on-one interviews to ensure an accurate understanding of the questions and reliable data recording. The survey was conducted via questionnaires, which mainly included the basic information of the surveyed elderly (including age, gender, height, weight, educational background, and presence or absence of chronic diseases, etc.) as well as the content of their outdoor activities within a day, activity duration (with 1 h as the basic unit), and the corresponding spaces where the activities occurred. A total of 193 valid questionnaires were retrieved (93 in summer and 100 in winter). The specific questionnaire content is presented in Appendix A.

2.4. Investigation on Outdoor Thermal Comfort

The outdoor thermal comfort survey of elderly residents was conducted during 21–27 July 2023, 22–24 August 2023, 7–13 January 2024, 24–26 January 2024, and 25–27 July 2024, covering nine villages: Nanqiangcun Village, Luojiazhuang Village, Taipinghe Village, Yezhai Village, Guocun Village, Shangcao Village, Xiacao Village, Daliang Village, and Songcun Village (Figure 4). The selection of these villages as survey sites was based mainly on climatic location, spatial morphology, and elderly activity scenes. All the surveyed villages are located in typical rural environments in the cold region of Xi’an and are representative of the winter and summer thermal environments in this area. Meanwhile, the nine villages share similar characteristics in terms of street and lane layout, residential form, and courtyard structure. Most dwellings are arranged as independent courtyard houses, representing a common residential spatial form in rural Xi’an. In addition, the daily activity spaces used by elderly residents in these villages include courtyards, doorways, lanes, and squares, thereby supporting a comprehensive analysis of the outdoor thermal environment of daily activity spaces for rural elderly residents.

The thermal comfort survey included subjective thermal comfort questionnaires and thermal environment measurements. All of the respondents were elderly people aged 60 and above with independent mobility (the sample selection method was consistent with that described in the previous section). The content of the subjective thermal comfort questionnaire covered two categories: first, the respondents’ basic information, including age, gender, height, weight, activity level in the 15 min prior to the survey, and clothing status; second, specific thermal comfort indicators, including thermal sensation (cold (−3), cool (−2), slightly cool (−1), neutral (0), slightly warm (+1), warm (+2), and hot (+3)) and thermal acceptability (completely unacceptable (−2), just unacceptable (−1), just acceptable (+1), and completely acceptable (+2)). While the questionnaires were being administered, measurements of thermal environment parameters—including air temperature, relative humidity, black globe temperature, and air velocity—were conducted using a heat index instrument (Portable Delta HD32.3, Delta Ohm S.r.l., Caselle di Selvazzano, Italy) (which meets the sensor measurement range and accuracy requirements specified in ISO 7726-2002 [43]) (The measuring range and accuracy of the instrument are shown in

Table 3. The measuring range and accuracy of the instruments.

Model	Measuring Range	Instrument Accuracy	Measurement
Heat index instrument (Portable Delta HD32.3)	Air temperature: −40–100 °C	1/3DIN	Investigation of outdoor thermal comfort/continuous monitoring of outdoor thermal environment.
	Relative humidity: 0–100%	±2%
	Black globe temperature: −30–120 °C	1/3DIN
	Air velocity:0.01–5 m/s	±(0.05 + 0.5% of reading) m/s
Temperature and humidity data logger (HOBOMX 2301)	Air temperature: −40–70 °C	±0.25 °C, (−40–0 °C) ±0.2 °C, (0–70 °C)	Continuous monitoring of outdoor thermal environment.
	Relative humidity: 0–100%	±2.5%
Black globe temperature self-recording instrument (HQZY-1)	Black globe temperature: −20–80 °C	±0.3 °C
Wind speed anemometer (Kestrel 5500)	Air velocity: 0.4–40 m/s	0.1 m/s

). The measurements were taken within a 1 m radius around each respondent, at a height level with their head. A total of 243 valid questionnaires were collected in this survey (113 in summer and 130 in winter). The specific questionnaire content is presented in Appendix B.

2.5. Continuous Monitoring of Outdoor Thermal Environment

Continuous monitoring of the outdoor thermal environment was conducted using three types of instruments: outdoor temperature and humidity data loggers (HOBO MX2301, Onset Computer Corporation, Bourne, MA, USA), black globe temperature recorders (HQZY-1, Beijing Tianjian Huayi Technology Development Co., Ltd., Beijing, China), and portable weather stations (Kestrel 5500, Nielsen-Kellerman Co., Boothwyn, PA, USA). The measuring ranges and accuracies of the instruments are shown in Table 3. These instruments were used to measure air temperature (t_a), relative humidity (RH), black globe temperature (t_g), and air velocity (v_a). Since the elderly mainly engage in seated activities when outdoors, the instruments were installed at a height of 1.1 m. Air temperature, relative humidity, black globe temperature, and air velocity were recorded because they are the basic environmental parameters required for UTCI calculation and thermal comfort evaluation. Specifically, black globe temperature and air velocity were used to calculate mean radiant temperature, and the measured wind speed was converted to the wind speed at 10 m height. The calculated UTCI values were then compared with the thermal comfort benchmark established from the thermal comfort field survey to evaluate the thermal comfort level of different outdoor activity spaces for older adults.

To avoid arbitrariness in the selection of survey locations and monitoring points, a daily activity survey was conducted before the formal thermal comfort field survey. In this stage, the occurrence time and corresponding outdoor activity spaces used by older adults’ daily activities on typical days were recorded to identify the outdoor spaces where older adults frequently stayed or gathered. Therefore, the monitoring points for the subsequent continuous thermal environment monitoring were further determined based on the results of the daily activity survey and the seasonal characteristics of older adults’ outdoor activity spaces. One typical rural residence in Songcun Village, Chang’an District, Xi’an, was selected for continuous thermal environment monitoring. This residence features a long and narrow courtyard measuring 24 m × 10 m, and its floor plan is shown in Figure 5. Based on the seasonal characteristics and types of elderly residents’ outdoor activity spaces, north-facing doorways and shaded areas of the courtyard were selected as monitoring points in summer, while south-facing sunny doorways and sunlit areas of the courtyard were selected as monitoring points in winter. The instruments were installed at these points to conduct continuous thermal environment monitoring. The monitoring periods were from 22 January to 28 January 2024 and from 22 July to 28 July 2024, corresponding to winter and summer, respectively. During the monitoring periods, the outdoor weather was predominantly sunny, which is the most frequent weather type in Xi’an during winter and summer.

3. Results and Analysis

3.1. Spatiotemporal Patterns of Outdoor Daily Activities

3.1.1. Content and Classification of Daily Activities

The daily activities among elderly residents in rural Xi’an during winter and summer were statistically analyzed. The results showed that the types of outdoor daily activities were generally consistent between winter and summer, with a total of 21 common activities identified (

Table 4. Activity modes and their classifications.

Daily Activity Content	Activity Type
Breakfast, lunch, dinner	Dining activities
Housework, sending and picking up children from school, childcare, caring for family members	Household activities
Sun exposure (in winter), sedentary rest, conversation	Leisure activities
Playing board games, mobile phone use, reading	Recreational activities
Performing farm work, growing vegetables, watering flowers, raising animals	Physical labor
Chinese square dance, cycling, massage, oxygen inhalation	Health-promoting physical activities

). For subsequent analysis, these activities were categorized into six types based on their behavioral characteristics and activity intensity: dining activities, household activities, leisure activities, recreational activities, physical labor, and health-promoting physical activities (HPPA).

The spatiotemporal differentiation characteristics of outdoor activities among elderly residents in rural Xi’an were statistically analyzed for winter and summer, respectively (Figure 6). The results indicated that, regardless of the season, leisure activities dominated the elderly’s outdoor routines, accounting for 65.80% in summer and 69.60% in winter. For physical labor, the proportion in winter (5.40%) was significantly lower than that in summer (17.20%). Among all activity types, outdoor dining activities accounted for the lowest proportion in summer, at 2.50%, whereas recreational activities accounted for the lowest proportion in winter, at 3.00%.

3.1.2. Temporal Distribution Patterns of Daily Activities

Statistical analysis was conducted based on the occurrence and duration of various outdoor activities among the elderly, with the results for summer and winter presented as follows:

• Summer

As shown in Figure 7a, the summer outdoor activities among elderly residents in rural Xi’an were mainly concentrated in the morning and evening periods, specifically from 6:00 to 9:00 and from 17:00 to 21:00. Dining activities were mainly concentrated during 7:00–8:00, 10:00–13:00, and 17:00–18:00. Household activities were more frequently distributed in the morning, especially from 8:00 to 11:00, accounting for approximately 52%. Leisure activities were widely distributed throughout the day and were mainly concentrated during 6:00–12:00 and 13:00–23:00. Recreational activities were mostly concentrated in the evening, from 19:00 to 21:00. Physical labor was more frequently distributed in the morning and afternoon. Health-promoting physical activities were mostly distributed from 7:00 to 10:00 in the morning and from 16:00 to 19:00 in the afternoon, accounting for approximately 74%. Overall, outdoor activities in summer were mainly concentrated during 6:00–9:00 and 17:00–21:00, and the elderly preferred to stay in shaded spaces.

2.

Winter

As shown in Figure 7b, the elderly’s outdoor activities in winter exhibited a “bi-modal” pattern, mainly concentrated during 9:00–11:00 and 13:00–17:00. Dining activities were mainly concentrated during 11:00–13:00, accounting for approximately 62%. Since temperatures are relatively low in the morning and evening in winter, outdoor dining activities were mostly conducted at noon. Household activities were distributed throughout the day, with higher frequencies in the morning, from 6:00 to 8:00, and in the early evening, from 16:00 to 17:00. Leisure activities were more widely distributed throughout the day and were mainly concentrated during 8:00–21:00. Recreational activities showed a relatively scattered distribution, with the highest frequency during 14:00–15:00, accounting for approximately 26% of total activities. Physical labor was mostly concentrated in the afternoon, from 15:00 to 16:00, accounting for approximately 23% of total activities. Health-promoting physical activities were more frequent in the morning, from 9:00 to 10:00, accounting for approximately 37%. Overall, in winter, the elderly’s outdoor activities were mainly concentrated during 9:00–11:00 and 13:00–17:00, and they preferred to bask in the sun in sunlit spaces.

3.1.3. Analysis and Discussion

Table 5. Comparison of elderly adults’ outdoor activities across regions.

City	Climate Zone	Region	Season	Activity Classification Method	Main Outdoor Activity Time	References
Xi’an	C	Rural	Summer, winter	Dining activities, household activities, leisure activities, recreational activities, physical labor, health-promoting physical activities	Summer: 6:00–9:00, 17:00–21:00 Winter: 9:00–11:00, 13:00–17:00	Current study
Xi’an	C	Urban	Autumn	Survival-oriented behavior, family responsibility-oriented behavior, life-oriented behavior	Three types of services are distributed throughout the day; Survival-oriented Behavior has no fixed travel time points. Family Responsibility-oriented Behavior is mostly distributed between 6:30–9:00, 10:30–11:15, and 13:30–17:00. Life-oriented Behavior is mostly distributed between 7:00–11:30, 14:30–17:30, and 18:40–19:50.	[44]
Ningbo	HSCW	Urban	Winter	Static activity, dynamic activity, group activity	10:00–11:00, 15:00–16:00	[24]
Beijing	C	Urban	Whole year	Work, housework, shopping, personal matters, sleep or naps, leisure and entertainment (intellectual type, emotional type, fitness type, communicative type, public welfare type)	6:00–10:00, 16:00–18:00	[45]
Shanghai	HSCW	Urban	Summer	Travel, Vegetable market shopping, Store shopping, Leisure and fitness, Cultural entertainment, Medical and health care, Social interaction	/	[46]

presents the temporal and spatial patterns of outdoor daily activities among the elderly across different studies. First, from the perspective of urban–rural differences within Xi’an, the distribution of activity time of the elderly in urban Xi’an exhibits stronger purposefulness than that of the elderly in rural Xi’an [44]. Further comparison between the elderly in rural Xi’an and those in urban areas of other climate zones or regions shows that the activity time and activity types of the elderly also vary across regions. The morning activity time of the elderly in Ningbo (climate zone: hot-summer and cold-winter zone (HSCW)) is later than that in rural Xi’an (climate zone: cold zone (C)) in winter, as documented in [24]. The activity time of the elderly in Beijing (climate zone: cold zone (C)) is earlier, more concentrated, and more regular than that of the elderly in rural Xi’an [45]. The types of activities among the elderly in the Shanghai area (climate zone: hot-summer and cold-winter zone (HSCW)) are more diverse than those of the elderly in rural Xi’an [46]. These comparisons indicate that the temporal and spatial patterns of outdoor daily activities among the elderly differ across urban–rural contexts, regions, and climate zones.

Overall, the activity time and activity types of the elderly differ across regions, and these differences are associated with climatic conditions, urban–rural infrastructure, socioeconomic status, and lifestyle differences. Morning activities start earlier in northern China, such as Beijing and Xi’an, and relatively later in southern China, such as Ningbo. The comparison shows that the interaction between environmental adaptability and lifestyles shapes the activity time and activity types of the elderly in different regions.

3.2. Thermal Environment, Thermal Response and Thermal Comfort

3.2.1. Outdoor Thermal Environment

The spatiotemporal patterns of elderly residents’ daily activities were analyzed to identify high-frequency outdoor activity spaces and typical activity periods. Based on this analysis, subjective thermal comfort surveys were conducted in these spaces during the corresponding periods. During the administration of the thermal comfort questionnaire to the elderly in rural Xi’an, the current state of the thermal environment where the elderly were located was simultaneously measured and recorded.

Table 6. Statistics of main parameters for outdoor thermal environment.

Season	Sample Size	Parameter	Minimum	Maximum	Mean	Standard Deviation
Summer	Male: 49 Female: 64	Air temperature ta/°C	27.00	36.30	31.87	2.09
		Relative humidity RH/%	28.40	74.60	55.43	11.84
		Air velocity va/m/s	0.00	1.43	0.37	0.26
		Black globe temperature tg/°C	28.70	38.80	32.29	1.91
		UTCI/°C	29.82	38.68	33.52	1.84
Winter	Male: 54 Female: 76	Air temperature ta/°C	3.20	16.00	8.45	3.57
		Relative humidity RH/%	13.90	77.20	42.80	18.95
		Air velocity va/m/s	0.03	2.37	0.39	0.35
		Black globe temperature tg/°C	4.00	20.50	10.09	4.20
		UTCI/°C	4.11	18.74	10.25	3.60

presents statistics on the minimum value, maximum value, average value, and standard deviation of parameters such as outdoor air temperature, relative humidity, air velocity, and black globe temperature during the summer and winter thermal comfort questionnaires, respectively. In this study on outdoor thermal environments, the UTCI index was used as the evaluation indicator. This index takes into account air temperature, mean radiant temperature, relative humidity, and air velocity at a height of 10 m. In this study, UTCI values were calculated using Rayman software (version 1.2, Meteorological Institute, University of Freiburg, Freiburg, Germany). The measurement height of the measured air velocity was 1 m above the ground surface, and the air velocity was corrected using the equations specified in ASHRAE [47]. The mean radiant temperature (t_mrt) was calculated in accordance with the ISO 7726 [43] standard.

3.2.2. Clothing Adjustment

Clothing adjustment is one of the most important measures for people to maintain thermal comfort. In the thermal comfort questionnaire, the clothing status of the elderly was recorded, and the overall clothing thermal resistance was calculated in accordance with ASHRAE 55 [48]. The thermal resistance of individual garments was determined with reference to the ASHRAE 55 standard. If the elderly were in a sitting posture, the overall clothing thermal resistance included the additional thermal resistance of the seat, and the calculation method for total thermal resistance followed the approach proposed by Mccullough [49]. As shown in

Table 7. Calculation of clothing thermal insulation (clo).

Season	Sample Size	Minimum	Maximum	Mean	Standard Deviation
Summer	Male: 49 Female: 64	0.338	0.720	0.463	0.072
Winter	Male: 54 Female: 76	1.190	2.232	1.690	0.200

, the average values of the clothing thermal resistance of the surveyed elderly were 0.4627 clo in summer and 1.6899 clo in winter.

To establish the pattern of how the elderly in rural Xi’an adapt to climatic conditions through clothing adjustment, we segmented UTCI values at 1 °C intervals, calculated and analyzed the average clothing thermal resistance and average UTCI value within each interval, and adopted linear regression for systematic statistical analysis. Statistical analysis was performed using IBM SPSS Statistics 29.0 (IBM Corp., Armonk, NY, USA), and all figures were plotted with OriginPro 2025 (OriginLab Corporation, Northampton, MA, USA). Correlation analysis revealed a significant correlation between clothing thermal resistance and UTCI. The fitted regression equation is shown as Equation (1), with an R² value of 0.96. The ANOVA results showed that the overall regression model was statistically significant (F = 518.163, p = 8.97 × 10⁻¹⁴ < 0.05), indicating that the fitted equation adequately represented the relationship between UTCI and clothing insulation, as shown in Figure 8.

$${\mathit{I}}_{\mathit{cl}}=-0.05018\mathit{UTCI}+2.17829$$

(1)

3.2.3. Thermal Sensation and Neutral Temperature

To study the subjective perception of environmental temperature among the rural elderly in Xi’an, we divided the UTCI values into 1 °C intervals, calculated the average UTCI value within each interval and the corresponding Mean Thermal Sensation Vote (MTSV), and adopted linear regression for statistical analysis to establish the correlation between UTCI and the thermal sensation among the rural elderly in Xi’an. Correlation analysis revealed a statistically significant relationship between UTCI and MTSV (Figure 9). Statistical analysis was performed using IBM SPSS Statistics 29.0 (IBM Corp., Armonk, NY, USA), and all figures were plotted with OriginPro 2025 (OriginLab Corporation, Northampton, MA, USA). The fitted regression equation is given by Equation (2), where R² is 0.72. The ANOVA results showed that the overall regression model was statistically significant (F = 41.419, p < 0.0001), indicating that the fitted equation adequately represented the relationship between UTCI and MTSV. The calculated outdoor neutral temperature was 10.19 °C, and the regression coefficient was 0.066; that is, for every 1 °C increase in outdoor UTCI, the thermal sensation increased by 0.066.

$$\mathit{MTSV}=0.066\mathit{UTCI}-0.67$$

(2)

The PET primarily reflects the physiological equivalent temperature that characterizes the comprehensive effects of the thermal environment [50]. The UTCI, based on the multi-node human heat transfer model and clothing thermal resistance model, enables more accurate simulation of human physiological responses to environmental thermal stress under complex climatic conditions [51].

Table 8. Comparison of elderly adults’ neutral temperature across regions (°C).

City	Climate Zone	Region	Population Type	Season	Activity Classification Method	Neutral Temperature	References
Xi’an	C	Rural	Elderly population	Summer, winter	MTSV = 0.066UTCI − 0.67 (R2 = 0.72)	10.19	Current study
Xi’an	C	Urban	Elderly population	Whole year	Whole year: MTSV = 0.0497PET − 0.6537 (R2 = 0.877) Summer: MTSV = 0.0678PET − 1.3785 (R2 = 0.639) Winter: MTSV = 0.0407PET − 0.5404 (R2 = 0.452)	Whole year: 13.20 (PET) Summer: 20.30 Winter: 13.30	[29]
Lhasa	C	Urban	Elderly population	Whole year	MTSV = 0.066PET − 1.360 (R2 = 0.781)	20.60	[52]
Lanzhou	C	Urban	Elderly population	Summer, autumn	Summer: MTSV = 0.10PET − 1.76 (R2 = 0.84) Autumn: MTSV = 0.14PET − 2.78 (R2 = 0.86)	Summer: 17.60 Autumn: 19.80	[32]
Dalian	SC	Urban	Elderly population	Summer	MTSV = 0.13PET − 2.93873	22.60 (PET)	[53]
Guangzhou	HSWW	Urban	Elderly population	Summer	MTSV = 0.276PET − 7.065 (R2 = 0.63)	25.60 (PET)	[30]

analyzes the neutral temperatures of the elderly in different thermal comfort studies. First, from the perspective of urban–rural differences within Xi’an, the neutral temperature among the elderly in rural Xi’an (climatic zone: C) is 10.19 °C, which is lower than that of the elderly in urban Xi’an (13.20 °C) [29]. Further comparison between the elderly in rural Xi’an and those in other regions shows that neutral temperature and thermal sensitivity also vary across regions and climate zones. The neutral temperature among the elderly in Lhasa (climatic zone: C) is 20.60 °C, which is higher than that of the elderly in rural Xi’an (10.19 °C) [52]. The sensitivity to temperature among the elderly in Lanzhou (climatic zone: C) is greater than that of the elderly in rural Xi’an (climatic zone: C) [32]. The elderly population in urban areas of Dalian (climatic zone: severe cold zone (SC)) demonstrates greater sensitivity to temperature variations than that in rural Xi’an (climatic zone: C) [53]. The elderly population in urban areas of Guangzhou (climatic zone: hot-summer warm-winter zone (HSWW)) exhibits the strongest sensitivity to temperature, and their thermal sensation is most strongly influenced by temperature [30].

Overall, these regions exhibit a neutral temperature gradient and variations in thermal sensitivity that range from humid–hot urban areas to dry–cold rural areas. Owing to factors such as differences in urban–rural living environments, variations in living conditions, and geographical altitude, the elderly in rural areas of Xi’an exhibit a lower neutral temperature. Compared to the elderly in rural Xi’an, those in Guangzhou and Dalian demonstrate greater sensitivity to temperature variations, which is associated with differences in their respective climate zones and geographical locations.

3.2.4. Thermal Acceptability and Acceptable Temperature Range

To study the outdoor thermal acceptability among the elderly in rural Xi’an, the summer and winter UTCI values were divided into 1 °C intervals using the temperature–frequency bin method, and the corresponding Percentage Dissatisfied (PD) was calculated. Quadratic polynomial regression was then used for subsequent statistical analysis. Statistical analysis was performed using IBM SPSS Statistics 29.0 (IBM Corp., Armonk, NY, USA), and all figures were plotted with OriginPro 2025 (OriginLab Corporation, Northampton, MA, USA). The relationship between outdoor UTCI and PD was established, as shown in Figure 10, yielding the fitted Equation (3) with an R² value of 0.69. The ANOVA results showed that the overall regression model was statistically significant (F = 18.555, p = 1.16 × 10⁻⁴ < 0.05), indicating that the fitted equation adequately represented the relationship between UTCI and PD. When PD = 10%, the 90% outdoor thermal acceptability range was 9.60–27.20 °C; when PD = 20%, the 80% outdoor thermal acceptability range was 6.20–30.60 °C.

$$\mathit{PD}=0.0014{\mathit{UTCI}}^2-0.052\mathit{UTCI}+0.466$$

(3)

Table 9. A comparison of the acceptable temperature ranges for the elderly across different regions (°C).

City	Climate Zone	Region	Population Type	Season	Activity Classification Method	Outdoor Thermal Acceptability Range (UTCI)	References
Xi’an	C	Rural	Elderly population	Summer, winter	PD = 0.0014UTCI2 − 0.052UTCI + 0.466 (R2 = 0.69)	90%: 9.60–27.20 80%: 6.20–30.60	Current study
Xi’an	C	Urban	Elderly population	Whole year	PD = 0.0749PET2 − 2.7592PET + 31.179 (R2 = 0.7603)	10.90–25.90 (PET) (90%)	[29]
Lanzhou	C	Urban	Elderly population	Summer, autumn	Summer: URV = 0.0012PET2 − 0.054PET + 0.57 (R2 = 0.86) Autumn: URV = 0.0018PET2 − 0.089PET + 1 (R2 = 0.80)	Summer: ≤33.00 Autumn: 15.00–33.00 (90%)	[32]
Chengdu	HSCW	Urban	Elderly population	Summer, autumn, winter	Summer: PD = 0.0012PET2 − 0.0437PET + 0.3442 (R2 = 0.6179) Autumn: PD = 0.0245PET2 − 0.8188PET + 6.81114 (R2 = 0.7064) Winter: PD = 0.0054PET2 − 0.1569PET + 1.1574 (R2 = 0.522)	Summer: ≤29.52 Autumn: 14.41–19.01 Winter: ≥10.62 (PET) (90%)	[54]
Dalian	SC	Urban	Elderly population	Summer	PD = 0.42396PET2 − 16.90794PET + 178.05524 (R2 = 0.9065)	≤27.08 °C (PET) (80%)	[53]
Guangzhou	HSWW	Urban	Elderly population	Summer	URV = 0.496PET2 − 26.667PET + 359.412 (R2 = 0.776)	≤31.15 °C (PET) (90%)	[30]

presents the outdoor acceptable temperature ranges for the elderly across different thermal comfort studies. First, from the perspective of urban–rural differences within Xi’an, the acceptable temperature range for the elderly in urban Xi’an (Climate Zone: C) is 10.9–25.9 °C, which is narrower than that for the elderly in rural Xi’an (9.6–27.2 °C) [29]. Further comparison between the elderly in urban Xi’an and those in urban areas of other climate zones or regions shows that the acceptable temperature ranges for the elderly also vary across regions. The upper limit of acceptable temperature for urban elderly residents in Lanzhou (climate zone: C) is 33 °C, which is higher than that for the elderly in urban Xi’an [32]. The lower limit of acceptable temperature for winter outdoor thermal comfort for the elderly in urban Chengdu (Climate Zone: HSCW) is 10.62 °C, which is close to that for the elderly in urban Xi’an, while the upper limit of acceptable temperature for summer outdoor thermal comfort is 29.52 °C, which is higher than that for the elderly in urban Xi’an [54]. The 80th percentile upper limit of acceptable temperature for the elderly in urban Dalian (Climate Zone: SC) during summer is 27.08 °C, which is slightly higher than that for the elderly in urban Xi’an, but lower than that for the elderly in rural Xi’an [53]. The upper limit of acceptable temperature for summer outdoor thermal comfort for the elderly in urban Guangzhou (Climate Zone: HSWW) is 31.15 °C, which is also higher than that for the elderly in urban Xi’an [30]. These comparisons indicate that the outdoor acceptable temperature ranges for the elderly differ across regions and climate zones.

Overall, the upper limit of acceptable temperature for the elderly in rural Xi’an is higher than that in urban areas because urban elderly residents have weaker adaptability to high temperatures due to differences in their living environments and lifestyles. For the elderly in other regions, such as Chengdu, Dalian, and Guangzhou, whose upper limits of acceptable temperature differ, this variation is associated with differences in their respective climate zones and urban–rural living habits. Even within the same climate zone, discrepancies in the elderly’s acceptable temperatures persist in Lanzhou, and these discrepancies are largely attributed to differences in climatic characteristics and living habits.

3.3. Evaluation of Thermal Environment in Outdoor Daily Activity Spaces

3.3.1. Continuous Testing of Outdoor Thermal Environment

By combining the timing of outdoor thermal environment tests, the observation time range was determined based on the duration of the elderly’s outdoor activities. In accordance with the technical specification that “the thermal environment measurement cycle shall be 24~48 h” in the Chinese national standard GB/T 50785 [55], to systematically characterize the thermal environment characteristics of outdoor spaces for the daily activities among the elderly in Xi’an in both winter and summer, this study selected typical local summer and winter periods, respectively, to carry out continuous monitoring of thermal environment parameters. The summer monitoring cycle was from 22 to 28 July 2024, during which 7 consecutive days of data acquisition were completed. After climatic representativeness verification against the multi-year historical average air temperature data for July in Xi’an, combined with the temporal rhythm of the elderly’s daily outdoor activities clarified in the previous first-stage survey, the observation data from 05:00 to 22:00 (17 h in duration) on 26 July 2024 (a typical sunny day) was finally selected as the core analysis sample for the summer thermal environment. The winter monitoring cycle was from 22 to 28 January 2025, during which synchronous 7-day continuous monitoring was carried out. After climatic representativeness verification against the multi-year historical average air temperature data for January in Xi’an, combined with the core outdoor activity period of the elderly in winter, the continuous observation data from 08:00 to 18:00 (10 h in duration) on a typical winter day within the monitoring cycle was finally selected as the representative analysis sample for the winter thermal environment. This lays the fundamental data support for the subsequent analysis of thermal environment differences between winter and summer and the research on age-friendly thermal environment optimization. Data collected on the same day as the outdoor tests were selected for analysis, and the selected test times are shown in

Table 10. Outdoor thermal environment testing time.

Season	Time	Cycle
Winter	24 January, 08:00–18:00	10 h
Summer	26 July, 05:00–22:00	17 h

:

1.

Summer

The results of continuous monitoring of the summer thermal environment are shown in

Table 11. Environmental measurement parameters of outdoor spaces in Summer.

Space	Parameter	Minimum	Maximum
Courtyard	Air temperature ta/°C	25.85	41.19
	Relative humidity RH/%	15.00	32.85
	Air velocity va/m/s	0.0	0.7
	Black globe temperature tg/°C	25.4	55.6
Entrance (north-facing)	Air temperature ta/°C	26.63	35.67
	Relative humidity RH/%	36.31	59.04
	Air velocity va/m/s	0.0	0.5
	Black globe temperature tg/°C	26.8	36.5

. Due to direct sunlight, the peak air temperature in the summer courtyard reached 41.19 °C, and the black globe temperature peaked at 55.6 °C. Both values were higher than those at the north-facing entrance shaded by buildings, where the air temperature ranged from 26.63 °C to 35.67 °C, and the peak black globe temperature was 36.5 °C. In terms of relative humidity, the north-facing entrance recorded a higher range, from 36.31% to 59.04%, compared with 15.00% to 32.85% in the courtyard. The air velocity at the two monitoring points showed little difference and was mainly concentrated within the range of 0–0.2 m/s.

2.

Winter

The results of continuous monitoring of the winter thermal environment are shown in

Table 12. Environmental Measurement Parameters of Outdoor Spaces in Winter.

Space	Parameter	Minimum	Maximum
Courtyard	Air temperature ta/°C	−3.61	5.40
	Relative humidity RH/%	15.00	33.67
	Air velocity va/m/s	0.0	0.7
	Black globe temperature tg/°C	−3.6	7.8
Entrance (south-facing)	Air temperature ta/°C	−3.55	11.03
	Relative humidity RH/%	15.00	53.96
	Air velocity va/m/s	0.0	1.2
	Black globe temperature tg/°C	−3.9	27.3

. Clear thermal differences were observed between the south-facing entrance and the courtyard space. Due to direct sunlight, the south-facing entrance had a peak air temperature of 11.03 °C, and a maximum black globe temperature of 27.3 °C. Both values were higher than those of the courtyard space without direct sunlight, where the air temperature ranged from −3.61 °C to 5.40 °C, and the black globe temperature ranged from −3.6 °C to 7.8 °C. The air velocity at the entrance ranged from 0 to 1.2 m/s, which was slightly higher than that in the courtyard, ranging from 0 to 0.7 m/s. This difference may be related to the building enclosure effect.

3.3.2. Evaluation of Outdoor Thermal Environment

• Summer

Figure 11 shows the continuous monitoring results of the thermal environment at the north-facing entrance and the courtyard in summer. The summer monitoring data indicated that the UTCI indices of the north-facing entrance (Figure 11a) and the courtyard (Figure 11b) both exhibited obvious diurnal variation characteristics and exceeded the acceptable temperature range, shown as the green-labeled area in the figure, during most time periods. However, the temperature fluctuation range of the courtyard was larger than that of the entrance.

Shielded by buildings from direct sunlight, the UTCI values of the north-facing entrance ranged from 27.39 °C to 36.19 °C, which were generally lower than those of the courtyard exposed to direct sunlight, ranging from 24.41 °C to 44.84 °C. Before 09:00, the UTCI values of the north-facing entrance fell within the 80% outdoor thermal acceptability range, 6.20–30.60 °C. After 09:00, the values exceeded the upper limit of 30.60 °C, indicating a decline in thermal environment quality.

For the summer courtyard, the UTCI values were within the 90% outdoor thermal acceptability range, 9.60–27.20 °C, before 08:30, and then fell within the 80% outdoor thermal acceptability range, 6.20–30.60 °C, between 08:30 and 10:00. From 10:00 to 20:00, the UTCI values continuously exceeded the upper limit of the 80% outdoor thermal acceptability range, 30.60 °C, indicating poor thermal environment quality. After 20:00, the UTCI values gradually returned to the 80% outdoor thermal acceptability range.

These results suggest that the courtyard space may face a higher heat stress risk during most daytime periods in summer. Therefore, sunshade measures should be considered to improve the summer thermal environment.

2.

Winter

Figure 12 presents the continuous monitoring results of the thermal environment at the south-facing entrance and the courtyard in winter. Owing to direct sunlight exposure, the UTCI values of the south-facing entrance, ranging from −3.40 °C to 22.00 °C, were higher than those of the courtyard without direct sunlight, ranging from −3.60 °C to 6.30 °C (Figure 12a,b). The minimum UTCI values at the two monitoring points were similar, whereas the maximum UTCI value at the south-facing entrance was higher. This difference may be mainly related to the solar radiation gain of the south-oriented space under the low solar altitude angle in winter.

The UTCI values of the south-facing entrance fell within the 80% outdoor thermal acceptability range, 6.20–30.60 °C, from 10:30 to 16:30, and reached the 90% outdoor thermal acceptability range, 9.60–27.20 °C, between 11:00 and 16:15. In contrast, the thermal environment quality of the courtyard remained relatively poor due to building shading. The overall UTCI range of the winter courtyard was −3.30–6.24 °C, and it only briefly met the 80% outdoor thermal acceptability range from 13:30 to 14:00.

During periods of poor thermal environment quality, the difference between the courtyard’s winter UTCI minimum value, −3.60 °C, and the lower limit of the 80% outdoor thermal acceptability range, 6.20 °C, was 9.80 °C. Similarly, the difference between the south-facing entrance’s UTCI minimum value and the same lower limit was 9.60 °C. These results suggest that both monitoring spaces may face cold stress risk in winter, and corresponding protective measures should be considered to improve thermal comfort. The temperature ranges and time periods corresponding to the outdoor 80% thermal acceptability range in summer and winter are presented in

Table 13. Evaluation results of outdoor thermal environment.

Season	Space	UTCI Range/°C	Outdoor 80% Thermal Acceptability Range/°C	Time Periods Corresponding to the 80% Outdoor Thermal Acceptability Range
Summer	Courtyard	24.40–44.80 °C	6.20–30.60 °C	Before 10:00; after 20:00
	entrance (north-facing)	27.40–36.20 °C	6.20–30.60 °C	Before 09:00
Winter	Courtyard	−3.60–6.20 °C	6.20–30.60 °C	13:30–14:00
	entrance (south-facing)	−3.40–22.00 °C	6.20–30.60 °C	10:30–16:30

.

3.4. Objective Improvement Strategies for Thermal Environment and Subjective Adaptation Suggestions for Humans

3.4.1. Priority Improvement Paths and Strategies for Thermal Environment

With the development of computer technology, in addition to traditional research on establishing Predicted Mean Vote (PMV) models, scholars have also begun to establish thermal comfort evaluation models using machine learning methods to study thermal comfort indicators.

In 2024, Wang [56] adopted machine learning algorithms (decision tree, logistic regression, support vector machine, and random forest) to establish a thermal comfort vote prediction model, and analyzed the impact of single-input parameters and multi-input parameters on the prediction performance of the prediction model. In 2025, Deng [57] developed an improved thermal comfort prediction model by integrating individual characteristic parameters such as height, weight, and gender. Using machine learning methods, they established an enhanced predictive thermal state model based on human physiological parameters and classified thermal comfort levels. The study assessed the degree of influence of various environmental parameters on thermal sensation, identified key influencing factors, and thereby determined the optimization path for the thermal environment. These studies indicate that machine learning methods can be used not only for thermal comfort prediction, but also for identifying the relative importance of factors affecting thermal sensation. Traditional statistical modelling methods, such as correlation analysis, multiple linear regression, logistic regression, and ordinal logistic regression, are commonly used to examine the relationships between thermal sensation and individual environmental or personal variables. However, these methods usually require predefined functional forms or assumptions about linear relationships among variables, and they may be sensitive to multicollinearity, data distribution, and interaction effects. Therefore, the random forest model was introduced as a complementary method to traditional statistical analysis because it does not require a prior assumption of linear relationships and can better capture nonlinear effects and interactions among variables.

Based on the random forest algorithm, this study is analyzed through the following steps.

Step 1: This study constructed a thermal sensation prediction model for the elderly, selecting six key parameters—air temperature, mean radiant temperature, relative humidity, air velocity, clothing thermal resistance, and metabolic rate—as input features.

Step 2: Data preprocessing was performed in this study. The data had been standardized in our previous research, and 243 valid data entries were retained.

Step 3: Based on the thermal comfort survey data, the original seven-point thermal sensation scale was recoded into a three-point numerical scale. Specifically, votes from −3 to −1 were merged into “cold” (−1), vote 0 was retained as “neutral” (0), and votes from 1 to 3 were merged into “hot” (1).

Step 4: During model construction, the feature dataset X was first established, with a total of M samples. Then, random sampling with replacements was performed on the preprocessed dataset using the Bootstrap method to generate multiple training subsets. A decision tree was constructed from each training subset, and the final random forest model was formed by integrating the results of all decision trees.

Step 5: To evaluate the predictive performance of the model, cross-validation and test-set evaluation were conducted. The cross-validation RMSE was 0.4091. On the test set, the model achieved an R² of 0.6492, an RMSE of 0.4289, and an MAE of 0.2772. These results indicate that the random forest regression model had acceptable predictive performance. Therefore, the model was further used to analyze the relative importance of each feature affecting older adults’ thermal sensation [58]. The model was not intended for deployment-oriented prediction but was mainly used to identify the relative importance of different factors affecting thermal sensation.

The weights of various influencing factors on the outdoor thermal sensation of the elderly were obtained by establishing a random forest model (

Table 14. Characteristic weights of various influencing factors for outdoor thermal sensation.

Serial Number	Feature	Weight
1	Mean radiant temperature tmrt/°C	0.232
2	Relative humidity RH/%	0.230
3	Air temperature ta/°C	0.186
4	Air velocity va/m/s	0.131
5	Clothing thermal resistance CTR/clo	0.125
6	Metabolic rate MR/Met	0.095

). Among these, the weights of the mean radiant temperature (0.232) and relative humidity (0.230) are the highest, indicating that these two indicators have the most significant impact on the outdoor thermal sensation of the elderly, followed by air temperature (0.186), which shows that it still plays an important role in thermal environment evaluation. In contrast, the weights of air velocity (0.131), clothing thermal resistance (0.125), and metabolic rate (0.095) are relatively lower, suggesting that these factors have a relatively secondary impact on the outdoor thermal sensation of the elderly but still make a certain contribution. From the perspective of weight distribution, the impact of environmental thermal conditions (mean radiant temperature, relative humidity, and air temperature) on the elderly’s thermal sensation is significantly higher than that of individual factors (clothing thermal resistance and metabolic rate), indicating that external climatic conditions play a dominant role in influencing the indicators of the elderly’s outdoor thermal sensation.

Through this study, it was found that climatic conditions have a significant impact on thermal sensation. In the optimization of the outdoor thermal environment, this study adopts the random forest algorithm to calculate the weight of each influencing factor and identify the impact degree and importance ranking of different factors on the outdoor thermal environment. On this basis, targeted regulation of key influencing factors can precisely meet the requirements of thermal environment control and further improve the pertinence and effectiveness of the optimization strategy.

Table 15. Improvement measures for different seasons.

Strategy Priority (Sorted by Weight)	Feature Parameters	Improvement Measures (Summer)	Improvement Measures (Winter)
1 (0.232)	Mean radiant temperature tmrt/°C	1. Providing shaded spaces (adding sunshade facilities, planting street trees, shade trees, etc.). 2. Low sky view factor (SVF) of building complexes (under no sunshade facilities: in summer, UTCI decreases with increased SVF) to reduce solar radiation penetration.	Increasing SVF.
2 (0.230)	Relative humidity RH/%	Green planting (plant transpiration).
3 (0.186)	Air temperature ta/°C	1. Increasing green coverage rate (to reduce temperature). 2. Rational green planting (including types of greening configuration, tree morphology, quantity, and distribution).	1. Using materials with low thermal conductivity as road materials. 2. Constructing solar greenhouses.
4 (0.131)	Air velocity va/m/s	1. Improving rural building layout (reserving main ventilation ducts and increasing inter-building spacing). 2. Aligning lane orientation with summer dominant wind direction to facilitate air inflow. 3. Planting ventilation-permeable deciduous tree species.	1. Planting evergreen tree species with large crown width (wind-shielding) and wind-shielding shrubs. 2. Avoiding lane orientation aligning with winter dominant wind direction. 3. Using front-row buildings to block cold winds.
5 (0.125)	Clothing thermal resistance CTR/clo	Reducing amount of clothing worn.	Increasing amount of clothing worn.
6 (0.095)	Metabolic rate MR/Met	Engaging in activities with low metabolic rate (e.g., walking and reading).	Engaging in activities with high metabolic rate (e.g., square dancing and fitness exercises).

lists the corresponding improvement strategies for the factors affecting outdoor thermal sensation. In the optimization of the outdoor thermal environment, the above strategies can be selectively applied in a targeted manner to achieve precise regulation of the outdoor thermal environment.

3.4.2. Suggestions for Individual Subjective Adaptive Behaviors

According to the research findings, the elderly’s outdoor activities in winter exhibited a “bimodal” pattern, mainly concentrated during 9:00–11:00 and 13:00–17:00, with a preference for south-facing sunlit areas. During 10:30–16:30, the UTCI values fell within the 80% outdoor thermal acceptability range. In summer, their outdoor activities were mainly concentrated in the morning and evening periods, specifically from 6:00 to 9:00 and from 17:00 to 21:00, and they tended to prefer north-facing shaded spaces. Based on these findings, temporal and spatial behavioral recommendations are proposed for the elderly to help rural elderly residents better adapt to changing climatic conditions and improve their quality of life.

• Adjustments to Activity Time and Activity Space. Based on the patterns in the elderly’s activity spaces, the UTCI of the south-facing doorways in winter fell within the 80% outdoor thermal acceptability range during 10:30–16:30, making this period suitable for outdoor activities, such as leisure activities and health-promoting physical activities. In summer, the UTCI in the courtyard fell within the 80% outdoor thermal acceptability range before 10:00 and after 20:00. Therefore, activities at the north-facing doorways are recommended before 9:00, while activities in the courtyard are recommended before 10:00 and after 20:00.
• Adjustments to the Content and Intensity of Daily Activities. According to the analysis of the patterns of the elderly’s daily activity types, their daily activities showed clear seasonal differences. In summer, outdoor activities were dominated by leisure activities and physical labor, while in winter, they were dominated by leisure activities and household activities. Due to the needs of agricultural production, rural elderly residents may need to engage in outdoor labor. Therefore, it is recommended that farm work be scheduled before 10:00 a.m. in summer to avoid high-temperature periods and reduce the risk of heat stress.
• Clothing Adjustment. Elderly individuals should make appropriate clothing adjustments, mainly by adding or removing garments, and should adjust their clothing reasonably in response to temperature changes. It is recommended that the elderly adopt a dynamic clothing adjustment mechanism and use a layered dressing strategy to cope with day–night temperature differences, while also ensuring protection against extreme temperatures.

3.5. Limitations and Future Work

The field studies in this research were mainly conducted in summer and winter, aiming to ensure the comfort and health of the elderly in the hottest and coldest environments throughout the year. For regions with four distinct seasons, such as Xi’an, the climate in spring and autumn is generally comfortable and pleasant. However, large temperature differences between morning and evening may still occur, and uncomfortable periods may also exist in outdoor activity spaces for the elderly.

In future studies, on the one hand, special field studies should be conducted in spring and autumn to establish more precise season-specific thermal response and thermal comfort evaluation models. This would help form continuous research outcomes with an annual cycle and further support the creation of a comfortable outdoor thermal environment for rural elderly under changing climatic conditions across different seasons. On the other hand, optimization design research on the factors related to the elderly’s thermal sensation should be conducted for thermally uncomfortable periods in different types of outdoor daily activity spaces. Specifically, measures such as optimizing plant configuration and adjusting building layout and orientation could be adopted to extend the duration of outdoor comfort, improve the level of outdoor thermal comfort, and enhance the thermal environment quality of various activity spaces.

4. Conclusions

This study focused on elderly individuals aged 60 years and above in rural Xi’an and conducted research on outdoor thermal environment evaluation and improvement based on the spatiotemporal patterns of their daily activities. By analyzing the content and spatiotemporal patterns of the elderly’s daily activities, this study constructed an outdoor thermal response model for the elderly, calculated thermal comfort evaluation criteria, evaluated the current status of the thermal environment in the elderly’s outdoor daily activity spaces, identified key factors influencing thermal sensation and their weights using the random forest algorithm, and proposed targeted outdoor environment improvement strategies and suggestions for human activity behavior. The specific conclusions are as follows:

• The outdoor daily activities among elderly residents in rural Xi’an exhibited clear seasonal spatiotemporal differentiation patterns. In summer, their outdoor activities were concentrated in the morning and evening periods, specifically from 6:00 to 9:00 and from 17:00 to 21:00, and the elderly preferred to stay in shaded spaces. Leisure activities were widely distributed, and health-promoting physical activities accounted for a high proportion. In winter, their outdoor activities were concentrated during 9:00–11:00 and 13:00–17:00, while they mainly engaged in indoor activities during other time periods. The elderly preferred sunny spaces, and various activities showed a relatively scattered distribution. Overall, the activity patterns were adapted to the climatic characteristics of summer and winter. Different activity types exhibited clear differences in spatiotemporal distribution, which may be related to regional and seasonal climatic characteristics and the living habits of urban and rural elderly residents.
• Elderly residents in rural Xi’an exhibited unique thermal response patterns and thermal comfort requirements in outdoor spaces. Their thermal response patterns differed from those of elderly individuals in urban Xi’an and other climate zones. The measured neutral temperature was 10.19 °C, and the 90% and 80% thermal acceptability ranges were 9.60–27.20 °C and 6.20–30.60 °C, respectively. These values also differed from those of elderly individuals in urban Xi’an, regions in the same climate zone but with distinct typical climatic characteristics, and other climate zones. These differences may be associated with climatic features, the living habits of urban and rural elderly residents, and disparities in the thermal environment quality of outdoor living spaces for urban and rural elderly residents.
• The outdoor daily activity spaces used by elderly residents in rural Xi’an exhibited different thermal environment comfort levels. In summer, the comfortable time periods of the courtyard, before 10:00 and after 20:00, were broader than those of the north-facing doorways before 9:00, but the daily UTCI peak in the courtyard was higher. In winter, due to direct sunlight, the UTCI acceptable time period of the south-facing doorways, from 10:30 to 16:30, was longer than that of the courtyard, from 13:30 to 14:00. An outdoor thermal sensation random forest model was established, and the results showed that the weights of environmental and individual parameters influencing thermal sensation, in descending order, were mean radiant temperature, relative humidity, air temperature, wind speed, clothing thermal resistance, and metabolic rate. Based on these results, targeted thermal environment improvement strategies for summer and winter were proposed, along with behavioral recommendations for adjusting activity time and space, as well as activity content and intensity.

In conclusion, this study innovatively introduces the research methods of time geography into thermal comfort and thermal environment research. It first identifies high-frequency outdoor activity spaces and typical activity periods, then establishes thermal comfort criteria, evaluates a typical outdoor space through a case study, and proposes priority improvement pathways for rural outdoor activity spaces in cold regions. In this way, this study forms a “time-segmented and zone-specific” method for precise thermal environment evaluation and priority improvement, based on the elderly’s daily activity contents and their dynamic spatiotemporal trajectories. These findings provide a behavior-oriented basis for sustainable age-friendly environmental design by helping planners and policymakers determine where, when, and how rural outdoor thermal environments should be improved.