Evidence-Based Design of Residential Outdoor Spaces Considering Age-Specific Activity Patterns and Microclimatic Conditions

Abstract

Urban residential outdoor spaces are increasingly affected by high temperatures, strong solar radiation, and uneven wind conditions, which influence residents’ outdoor activities and thermal comfort. This study proposes an evidence-based approach for the fine-grained design of residential outdoor activity spaces based on age-specific activity patterns and microclimatic conditions. Using 1096 valid questionnaires, field observations, and in situ microclimate measurements across four seasons, we quantified the activity patterns, temporal distributions, and spatial preferences of children, adolescents, adults, and older adults. Results reveal statistically significant age- and season-dependent differences in environmental preferences (Kruskal–Wallis H = 56.78, p < 0.001 for light priority; H = 26.81, p < 0.001 for thermal priority across age groups). Children and older adults exhibited sustained and widely distributed activities, whereas adolescents and adults showed more concentrated temporal patterns. In summer, activities shifted toward mornings and evenings to avoid heat, while in winter, activities peaked around midday and afternoon. Wind and light conditions were prioritized over thermal conditions across all seasons: in summer, 47.2% of respondents ranked wind first; in winter, 50.9% ranked light first. Tree shading consistently reduced air temperature, black globe temperature, and WBGT relative to open areas in every season, with the largest mean differences observed in spring (ΔT_g = 1.19 °C) and the largest instantaneous difference during summer early afternoon (ΔT_g = 1.51 °C at 13:00). Integrating these findings, this study proposes evidence-based design strategies, including optimized functional layouts, shading and ventilation features, sun-exposure management, and nighttime lighting, providing quantitative support for improving comfort, safety, and usability in residential outdoor spaces.

1. Introduction

With global climate change and increasing urban density, outdoor thermal environments in residential neighborhoods have become a critical public health challenge. Rising heat exposure constrains residents’ outdoor activity and well-being through several interconnected pathways: it limits the performance metrics by which climate-conscious urban design is judged, reduces children’s opportunities for active outdoor play, and intensifies heat stress in hot and humid regions where integrated space design is most needed [1,2,3]. The consequences extend beyond physical activity to less visible outcomes, including mental work performance in subtropical climates and the social role that shared outdoor settings play in community life, for instance in supporting immigrant integration [4,5]. Current design practices often rely on aesthetic judgment and experience rather than quantitative assessments of key environmental factors such as wind, sunlight, and thermal conditions. Moreover, they frequently overlook substantial differences in physiological needs, behaviors, and environmental preferences across age groups, resulting in underutilized spaces, heat stress in summer, and wind exposure in winter, particularly among vulnerable populations such as children and older adults.

A growing body of field studies has demonstrated that outdoor microclimatic conditions—particularly solar radiation intensity and wind speed—exert significant spatial and temporal effects on pedestrians’ immediate thermal responses, yet these studies also reveal how strongly the monitoring approach itself shapes the conclusions that can be drawn. Xie et al. [6] used a backpack-mounted microclimate station carried alongside pedestrians walking a 1.2 km route in Beijing, which allowed solar intensity and wind speed to be resolved at the spatial resolution of individual stopping points; this mobile design is precisely what allowed them to detect that thermal sensation at any given point depends on the preceding 20–35 s of thermal history, a transient effect that a fixed-station or single-point design would not be able to observe. Earlier mobile and longitudinal monitoring work reached a related conclusion by different routes: Lau et al. [7] showed that pedestrian thermal comfort under outdoor transient conditions diverges systematically from steady-state predictions, while Krüger et al. [8] found that the relevant acclimatization window itself varies with exposure duration and season, ranging from short-term within-day adjustment to longer heatwave-driven adaptation. Together, these studies indicate that any monitoring protocol collapsing exposure into a single time-averaged reading risks masking the very temporal dynamics that govern how people actually experience outdoor space, a methodological concern directly relevant to the present study’s own continuous, site-resident microclimate monitoring approach (Section 2.2).

A second, largely independent methodological challenge concerns not when outdoor conditions are sampled but how the radiant environment is represented once sampled. We acknowledge this limitation explicitly because it bears directly on the globe-thermometer-based measurements used in this study: mean radiant temperature (T_mrt) is the variable most often singled out in this respect, with multiple independent studies documenting that the globe thermometer—though widely used outdoors for its low cost and ease of mobile deployment—is geometrically and operationally limited. Kántor and Unger [9] characterized Tmrt as the most problematic variable in human-biometeorological comfort assessment precisely because no single low-cost instrument measures it directly; every field deployment must therefore choose a proxy, and that choice carries its own bias. Tang et al. [10] showed in a subtropical field comparison that different observation methods for outdoor T_mrt can disagree substantially under the same sky conditions, and Lamberti et al. [11] traced this disagreement to instrument geometry itself: by comparing radiative fluxes computed for a sphere against those for a cylinder approximating a standing pedestrian across open, semi-open, and canyon urban morphologies, they showed that the spherical shape systematically overestimates diffuse shortwave radiation absorption and underestimates the directional, view-factor-dependent exposure a standing body actually receives, leading to T_mrt overestimation—and consequently overestimation of derived stress indices such as Physiologically Equivalent Temperature (PET) and Universal Thermal Climate Index (UTCI)—that is largest in sun-exposed open areas with high sky view factors and comparatively small in shaded or canyon settings. Read together, these three studies [9,10,11] indicate that globe-based T_mrt estimates are not a uniform source of error to be corrected by a single offset, but a context-dependent bias whose magnitude depends on sky exposure, urban morphology, and time of day. This has a direct bearing on field campaigns, including the present study, that rely on compact weather-station globe thermometers for black globe temperature (T_g) and derived Wet-Bulb Globe Temperature (WBGT) measurements: such results should be interpreted as indicative of relative differences between sites (e.g., shaded versus open) rather than as absolute, instrument-independent estimates of radiant exposure, a distinction we return to when interpreting our own shading comparisons (Section 3.2 and Section 4).

These two methodological strands—the temporal resolution at which exposure is sampled, and the geometric fidelity with which the radiant environment is represented—are seldom discussed together, yet both ultimately limit how confidently field-measured microclimate data can be linked to people’s reported thermal experience and behavior. The present study’s combination of continuous on-site monitoring with questionnaire- and observation-based behavioral data is one way of partially compensating for these limitations, since behavioral and preference outcomes are not solely dependent on the absolute accuracy of any single instantaneous physical measurement; nonetheless, the measurement-related caveats above apply to the microclimate component of this study and are revisited in Section 4. These findings, taken together, highlight the complex, transient nature of outdoor thermal exposure and the limitations of relying solely on steady-state indices or single-geometry instruments for behavioral prediction.

Although environmental exposure has been shown to profoundly impact health, most studies focus on single dimensions, limiting their practical application in comprehensive design strategies. Heat stress, in particular, poses heterogeneous risks across age groups, a heterogeneity that has been documented from several angles. Among older adults, age-related differences in thermoregulation and adaptive capacity make outdoor thermal stress a particular concern: the key factors shaping their outdoor thermal comfort have been shown to vary across climate zones, with gender-specific seasonal differences also reported, and these patterns have in turn informed outdoor thermal benchmarks for nursing-home design [12,13,14]. Related vulnerable-population concerns extend to clinical settings, where outdoor thermal comfort has similarly informed hospital open-space design [15]. Children represent a second vulnerable group: because of their higher activity intensity and still-developing thermoregulatory capacity, outdoor environmental conditions in kindergartens have been shown to significantly influence children’s moderate-to-vigorous physical activity, with diurnal differences also observed in their thermal comfort and adaptive behavior during outdoor play in subtropical regions [16,17]; portable shading interventions, in turn, can meaningfully shift pedestrians’ thermal sensation and adaptive behavior under strong solar exposure [18]. A second strand of evidence concerns the modeling approach itself rather than the population: traditional steady-state thermal comfort models are limited in explaining real outdoor behavior, since human thermal responses differ substantially under step changes between indoor and outdoor conditions compared with steady-state assumptions—a limitation that has prompted dedicated design-strategy reviews for the transient conditions characteristic of semi-outdoor spaces [19,20]. These findings indicate that residential outdoor spaces should not be evaluated only by generalized thermal indices, but should also consider age-specific activity rhythms, behavioral adaptation, and seasonal environmental preferences.

Parametric design and numerical simulation tools, such as the three-dimensional microclimate simulation software ENVI-met and computational fluid dynamics (CFD), offer support for microclimate optimization, and have been applied at two complementary scales. At the configuration scale, multi-objective optimization of urban block layout and parametric optimization of residential morphology have both been used to enhance outdoor thermal comfort, in the latter case while also accounting for cooling performance and site development constraints [21,22]. More recent work has pushed these tools toward multi-criteria, data-driven design: multi-phase frameworks now balance building energy consumption, solar energy potential, and outdoor thermal comfort within residential blocks; parametric modeling combined with neural networks has been used to generate and evaluate residential layouts against sky view factor, sunshine duration, and noise, explicitly targeting the incompleteness of single-indicator evaluation systems; and comparative optimization has been applied across high-rise versus multi-storey residential layouts [23,24,25]. A further strand of green-space optimization research argues that vegetation layout, daylighting, thermal comfort, and green view should be coordinated jointly rather than optimized as isolated indicators—an argument supported by frameworks that jointly optimize residential green space for outdoor thermal comfort, indoor daylight, and Green View Index, and by design-optimization frameworks that jointly evaluate building energy performance and outdoor thermal comfort across urban typologies [26,27]. However, existing approaches typically focus on universal physical indicators and morphological parameters and still lack a clear pathway to translate population demand maps into actionable design parameters. The applicability of such models across diverse climate zones and age-specific activity scenarios also requires further validation.

To address these gaps, this study proposes an evidence-based method for fine-grained design of residential outdoor activity spaces, integrating age-specific activity patterns with wind–thermal environmental evaluation. By combining surveys, field observations, and microclimate measurements, we quantify activity patterns and spatial preferences of children, adolescents, adults, and older adults, and analyze how wind, light, and thermal conditions influence residents’ spatial choices and activity patterns. Here, the “light environment” refers primarily to residents’ perceived sunlight exposure and shading conditions rather than measured illuminance. Across-season analyses reveal age-specific tendencies and priorities in space selection under varying environmental conditions. This approach enables construction of a multidimensional evaluation framework that informs population-sensitive design strategies, including shading, ventilation corridors, sun-exposure management, and nighttime lighting, providing a practical reference for improving comfort, health, and usability of residential outdoor spaces.

2. Materials and Methods

2.1. Study Sites

This study was conducted at typical residential outdoor activity sites in Guangzhou, China. Guangzhou is located in a subtropical monsoon climate zone, characterized by hot and humid summers and relatively cold and dry winters, resulting in pronounced seasonal variations in outdoor thermal conditions, which provide a representative climatic context for this research [12,17]. Six representative residential outdoor activity sites (denoted A–F) were selected, covering a variety of functional spaces, including children’s play areas, fitness zones, leisure areas, open plazas, and sports courts, to capture how different spatial functions influence human activity patterns and microclimate perception.

To characterize microclimate differences under varying shading and ventilation conditions, paired measurement points were arranged where site conditions allowed. Sites A, B, E, and F included complete open-shaded paired points, whereas sites C and D were dominated by tree-shaded areas and were therefore used mainly for activity observation and spatial preference analysis. The locations and basic characteristics of all surveyed sites are summarized in

Table 1. Detailed Information of Surveyed Residential Outdoor Activity Sites.

Site ID	Site Dimensions, Length × Width (m)	Site Openness and Shading Conditions	Distribution Map of Site Measurement Points	Site Photographs
A	28.0 m × 12.5 m	Open unshaded area + tree-shaded area
B	50.0 m × 21.0 m	Open unshaded area + tree-shaded area
C	24.5 m × 12.0 m	Tree-shaded area
D	34.0 m × 14.5 m	Tree-shaded area
E	42.0 m × 14.5 m	Open unshaded area + tree-shaded area
F	66.0 m × 33.0 m	Open unshaded area + tree-shaded area

. Using typical meteorological year data for Guangzhou from the Chinese building standards weather database, the Physiologically Equivalent Temperature (PET) was calculated to define seasonal boundaries, and the year was divided into four seasons: winter (early December to late February), spring (early March to late April), summer (early May to mid-October), and autumn (mid-October to late November). Field surveys were conducted on clear and favorable days for outdoor activities, covering both weekdays and weekends, and synchronized on-site and online questionnaires were distributed during each survey period.

Due to the geographic distribution of the survey sites, sites A, B, C, D, and E, which are within approximately 1 km of each other, were measured simultaneously using mobile instruments, while site F was measured at fixed points on days with weather conditions similar to the other five sites to ensure data comparability. The layout of measurement points at each site is detailed in Table 1, ensuring representative microclimate data collection across different functional zones, shading conditions, and ventilation environments, providing a reliable foundation for subsequent analysis of population activity patterns and environmental perception rankings.

Typical meteorological year (TMY) data for Guangzhou were used to calculate the PET for the purpose of defining seasonal boundaries. Field surveys were conducted on clear days, covering both weekdays and weekends.

2.2. Microclimate Measurement and Calibration

To accurately characterize the microclimate conditions at each site, this study employed the Kestrel 5400 handheld weather meter (Nielsen-Kellerman Co., Boothwyn, PA, USA), which simultaneously measures air temperature (T_a), relative humidity (RH), wind speed (V_a), and T_g. WBGT is calculated by the instrument from measured inputs rather than directly measured. These parameters are commonly used to characterize outdoor microclimate conditions and human thermal perception [1,6]. Measured parameters and instrument accuracy are summarized in

Table 2. Measured Parameters and Instrument Accuracy.

Parameter	Measurement Range	Accuracy	Resolution	Sampling Frequency
Ta	–29–70 °C	0.5 °C	0.1 °C	10 s
RH	10–90%	±2%	0.1%
Va	0.6–40 m/s	±3%	0.1 m/s
Tg	–29–60 °C	1.4 °C	0.1 °C

.

To verify the measurement accuracy and precision of the Kestrel 5400 and to determine the stabilization time of its data under mobile measurement conditions, the device was compared against validated fixed-point instruments, namely the HOBO temperature and humidity data logger (Onset Computer Corporation, Bourne, MA, USA) and the HD32.3 globe thermometer (Delta Ohm S.r.l., Caselle di Selvazzano, Padova, Italy) . Calibration experiments were conducted under clear weather conditions, with simultaneous measurements in both open unshaded areas and tree-shaded areas. The results indicated that the Kestrel 5400 data trends were consistent with the reference instruments. T_g and WBGT readings were slightly higher than those of the fixed instruments, whereas RH and V_a measurements were comparable.

To minimize instrument-specific errors and improve data reliability, the measurement protocol was optimized by extending the stabilization time. Measurements were compared following different waiting times (2, 4, 6, 8, and 10 min) after powering on the instrument at the measurement point. The analysis showed that waiting for 8 min before recording, and then averaging the data collected between the 9th and 13th minute, yielded values most consistent with the reference instruments. Therefore, in this study, the final protocol involved powering on the Kestrel 5400 at each site, waiting 8 min, and recording the 5 min average from minutes 9 to 13 as the representative microclimate parameters for that period. This procedure constitutes instrument acclimatization rather than formal calibration in the strict metrological sense, as transient outdoor conditions preclude the controlled reference environment that calibration would require. The parameters of these reference instruments are summarized in

Table 3. Parameters of Reference Instruments for Calibration.

Surveyed Measurement Project	Measurement Parameter	Measurement Range	Accuracy	Sampling Frequency
HOBO X100–011A Data Logger (placed in Stevenson screen)	Ta	−20–70 °C	±0.21 °C	1 min
	RH	1–95%	±2%
HD32.3 Thermal Index Instrument	Va	0.05–5 m/s	±0.05 m/s (0–0.99 m/s), ±0.15 m/s (1–5 m/s)
	Tg	−10–100 °C	±0.2 °C

. Photographs of the surveyed instruments and reference instruments deployed in the field are shown in Figure 1.

2.3. Observation of Population Activities

To capture activity patterns across different age groups in residential outdoor spaces, this study conducted field observations combined with photographic records. Observed participants included children (0–6 years), adolescents (7–17 years), adults (18–50 years), and older adults (>50 years), covering both genders and a range of activity preferences. Observations were carried out at six representative sites (A–F), including children’s play areas, senior activity zones, and public spaces. Each site was monitored across four seasons: spring, summer, autumn, and winter, and three daily time periods: morning, afternoon, and evening.

Activity types were classified according to intensity and form, including stationary activities, low-intensity activities, and moderate-to-high-intensity activities. This classification helps link observed behavior with activity intensity and spatial environmental characteristics [2,16]. Each observation record included a timestamp, activity type, number of participants, spatial location, and duration. These data were compiled into a structured database to quantify activity patterns across sites, time periods, and age groups. In addition to structured questionnaire items, brief informal interviews were conducted with a subset of on-site residents during field visits, in which several respondents explicitly expressed a desire for improved ventilation at their activity sites; this qualitative feedback, together with the structured wind-preference data (Q11) and observed seasonal shifts toward ventilated spaces, informed the ventilation corridor recommendation discussed in Section 4.

2.4. Questionnaire Survey

To collect residents’ subjective perceptions and behavioral preferences regarding residential outdoor activity spaces, this study employed a combined online and offline questionnaire survey. For children aged 0–6 years, offline questionnaires were completed by accompanying guardians to ensure data completeness and accuracy. Across the four seasonal surveys, a total of 1149 questionnaires were collected, of which 53 were excluded as invalid, yielding 1096 valid responses and an overall valid response rate of 95.3%. Specifically, 279 valid responses were obtained in spring, 284 in summer, 266 in autumn, and 267 in winter. The seasonal distribution and demographic characteristics of valid respondents are summarized in

Table 4. Seasonal Distribution and Demographic Characteristics of Valid Respondents.

Age Group	Spring	Summer	Autumn	Winter	Total
Children (0–6 yr)	40	34	28	37	139
Adolescents (7–17 yr)	38	45	33	26	142
Adults (18–50 yr)	130	142	137	136	545
Elderly (>50 yr)	71	63	66	68	268
Total	279	284	266	267	1096

. For the environmental priority-ranking item (Q13) specifically, two autumn-wave responses were missing or unparseable; the effective autumn sample for all priority-ranking analyses (Section 3.3) is therefore n = 264. The invalid response rate ranged from 4% to 6%, and the gender distribution was relatively balanced across seasons. After merging the detailed age categories into the four analytical groups used in this study, the sample covered children, adolescents, adults, and older adults across all seasons, providing a basis for analyzing age-specific activity patterns and environmental preferences.

The questionnaire included the following sections:

• Basic Information: Age group (0–6 years, 7–17 years, 18–50 years, >50 years) and gender.
• Activity Time: Residents’ outdoor activity periods across spring, summer, autumn, and winter, distinguishing weekdays from weekends, covering the full–day cycle from 6:00–7:00 to 21:00–22:00.
• Activity Types: Main outdoor activities by season and time, such as socializing, playing cards, walking, and fitness.
• Environmental Perception and Preference: Key environmental constraints affecting outdoor activity, including wind, sunlight exposure, shading, and temperature, were investigated. Residents’ preferences were categorized in terms of light environment conditions (sunny, partial shade, shaded), wind conditions (no wind, light breeze, wind corridor), and thermal conditions (warm, moderate, cool). In this study, the light environment mainly refers to residents’ perceived sunlight exposure and shading conditions rather than measured illuminance. Participants were asked to rank the priority of these environmental factors for the locations they frequent, allowing quantification of the relative importance of environmental conditions for different population groups [1,6].
• Spatial Selection Preference: Seasonal preferences for different types of outdoor activity spaces, such as children’s playgrounds with slides or senior leisure areas with pavilions.

Nighttime lighting was assessed via a dedicated item asking whether improved outdoor lighting would increase the respondent’s evening activity duration. Among all respondents, 25.3% indicated that improved nighttime lighting would extend their outdoor activity time. The full questionnaire instrument is provided as Appendix A.

The survey data were used to analyze seasonal and temporal activity characteristics, environmental preferences, and spatial selection tendencies across age groups, providing a quantitative basis for subsequent activity pattern analysis and fine-grained space design.

2.5. Data Analysis and Design Translation

Questionnaire data were screened and classified by age group, season, activity time, activity type, and environmental preference. Descriptive statistics were used to identify temporal activity patterns and preference distributions across different population groups. Non-parametric Kruskal–Wallis tests were applied to ordinal priority ranking data (scale: 1 = most important, 3 = least important) to assess differences across age groups and seasons. Pearson chi-square (χ²) tests were used to assess categorical differences in the proportion of respondents selecting each environmental factor as their top priority. Where omnibus tests reached significance (p < 0.05), post hoc pairwise comparisons were conducted using Dunn’s test with Bonferroni correction. Statistical analyses were performed in Python 3.12 (Python Software Foundation, Wilmington, DE, USA) with SciPy v1.11 (SciPy Developers, Austin, TX, USA). Field observation records were organized to construct activity time-space distribution maps, allowing comparison of activity intensity and spatial use among seasons and age groups. Microclimate measurements were compared between open unshaded and tree-shaded areas to quantify seasonal differences in T_a, RH, V_a, T_g, and WBGT. The questionnaire, observation, and microclimate results were then integrated to identify key behavior-environment relationships. Based on these relationships, design strategies were derived for functional zoning, shading provision, ventilation organization, sun-exposure management, and nighttime lighting.

3. Results

3.1. Seasonal and Age-Specific Activity Patterns and Spatial Selection

During synchronous microclimate measurements in residential outdoor spaces, this study recorded outdoor activity types and spatial choices of residents across six representative sites (A–F). Observed participants were classified as children (0–6 years, requiring supervision), adolescents (7–17 years, able to act independently but constrained by school schedules), adults (18–50 years, with activities concentrated during weekends and weekday evenings due to work pressure), and older adults (>50 years, primarily engaged in fitness and social activities without strict time constraints). Data were analyzed by season and across the daily period from 6:00 to 21:00, combined with questionnaire results, to construct activity time–space distribution maps for each age group. These maps visually illustrate sustained and widespread activity for children and older adults, concentrated activity for adolescents and adults, and seasonal adjustments in activity patterns (Figure 2).

Observations indicate that while activity types and spatial distributions are generally stable across age groups, pronounced seasonal adjustments exist. Children’s activities were mainly concentrated in playground areas, with high-intensity activities (running, climbing, slides) typically occurring between 8:00–11:00 and 15:00–18:00. Activity decreased during summer peak heat (11:00–15:00) and shifted in winter to 10:00–12:00 and 14:00–17:00, with evening activity shortened to 19:00–20:00. In summer, children’s peak activity window contracted by approximately 3 h relative to spring (to 08:00–10:00 and 16:00–19:00), reflecting behavioral thermoregulation. Adolescents exhibited highly concentrated activity patterns, primarily during weekday mornings and evenings (7:00–19:30) and weekend afternoons (14:00–17:00), engaging mainly in badminton, roller-skating, and social activities. Summer activity shifted to early morning and evening in shaded areas due to high temperatures, while winter activity favored sun-exposed walkways and plazas. Adults mainly engaged in low-intensity activities (walking, jogging, fitness), peaking on weekday evenings (17:30–20:30) and weekends (9:00–11:30 and 15:00–17:30). Summer activity utilized shaded paths and plazas, whereas winter activity preferred sunlit areas. Older adults maintained relatively continuous activity throughout the year, mainly in pavilions, resting areas, and fitness zones for card games, socializing, Tai Chi, and dancing, with activity periods lasting up to 2–3 h per session. Some older adults synchronized their activities with children in playgrounds or open plazas, sharing peak activity periods with children.

Facility type exerted pronounced age-specific effects on spatial selection. Children’s activity was concentrated in playground areas with accompanying supervision, while pavilions and resting areas attracted older adults for socializing and relaxation. Fitness equipment areas maintained basic participation across all age groups. Nighttime lighting extended usable activity time, with well-lit areas showing approximately 30% more children engaged in activities between 18:00 and 21:00 compared to poorly lit areas, consistent with questionnaire findings that 25.3% of respondents reported improved lighting would extend their evening outdoor time. Site location indirectly influenced activity scale and age composition through “opportunity for stay–age adaptability”; internal community sites attracted a greater diversity of age groups, while peripheral sites exhibited more concentrated activity numbers and age distributions.

Overall analysis indicates that residential activity patterns are governed by a composite mechanism of “time-constraint driven + facility/environmental perceptibility regulation + location-based amplification”: adolescents and adults’ activity rhythms were mainly driven by work or study schedules, children’s activities were dominated by supervision requirements, microclimate factors guided fine-scale spatial selection, facility types shaped age-specific preferences, nighttime lighting extended usable time, and site location modulated overall activity scale and diversity. This pattern is broadly consistent with previous findings that outdoor activity is jointly affected by environmental exposure, spatial features, natural elements, and age-related thermal vulnerability [2,16,17].

3.2. Instrument Acclimatization and Microclimate Measurement Results

Instrument acclimatization tests showed that the Kestrel 5400 stabilized after approximately 6 min, and measurements recorded after 8 min showed the closest agreement with the reference instruments. Therefore, the 8 min stabilization protocol was adopted for all field measurements. The acclimatization differences under different waiting times are summarized in

Table 5. Measurement Differences in Kestrel 5400 Weather Meter under Different Waiting Times.

Waiting Time	Measurement Difference
	Ta (°C)		RH (%)		Va (m/s)		Tg (°C)		WBGT (°C)
	Min	Max	Min	Max	Min	Max	Min	Max	Min	Max
2 min	4.27	6.55	11.72	21.98	−0.29	0.20	4.88	6.91	1.50	2.04
4 min	1.61	4.57	5.26	15.32	−0.39	0.20	3.13	5.46	0.83	1.46
6 min	0.84	2.45	−1.00	3.46	−0.28	0.08	1.41	2.81	0.46	0.72
8 min	0.75	2.19	−0.89	0.81	−0.15	0.04	1.54	2.72	0.45	0.76
10 min	0.86	2.17	−0.55	0.79	−0.16	0.06	1.80	2.76	0.46	0.75

. This procedure constitutes acclimatization—ensuring instrument thermal equilibrium with the environment—rather than formal calibration in the strict metrological sense, which would require a controlled reference environment not achievable outdoors.

Following instrument acclimatization, this study conducted comparative measurements of microclimate variations across different spatial configurations in residential sites, focusing on three key outdoor environmental dimensions: wind, light, and thermal conditions. Considering that sites C and D lacked a complete open–shaded area comparison, seasonal analyses were conducted at sites A, B, E, and F, which possessed fully paired open unshaded and tree-shaded areas. Results indicated that tree-shaded zones consistently improved thermal conditions across all seasons, and the cooling effect was more pronounced during periods with stronger solar exposure, consistent with previous studies showing that vegetation coverage, canopy structure, and landscape configuration can regulate radiation exposure and improve local outdoor thermal comfort [3,20,27,28].

Seasonal differences are summarized in

Table 6. Seasonal Differences in Key Microclimate Parameters Between Open Unshaded and Tree-Shaded Areas.

Season	ΔTa * (°C)	ΔRH (%)	ΔTg (°C)	ΔWBGT (°C)	Main Difference Period	Observed Effect
Spring	0.95	2.00	1.19	0.30	11:00–15:00	Shading shows stable cooling and radiation reduction
Summer	0.75	2.86	0.85	0.42	12:00–14:00	Thermal improvement most pronounced; shading effect strongest
Autumn	0.55	3.37	0.64	0.47	12:00–14:00, 19:00–21:00	Continued significant thermal reduction
Winter	0.50	3.39	0.58	0.50	9:00–14:00	Improvement present, mainly during daytime

* Δ values represent “open unshaded area—tree-shaded area”; positive values indicate that the open area is warmer than the tree-shaded area, i.e., shading provides a cooling benefit.

. Mean differences (open minus shaded, averaged across the full daily measurement period) were: spring, ΔT_a = 0.95 °C, ΔT_g = 1.19 °C, ΔWBGT = 0.30 °C; summer, ΔT_a = 0.75 °C, ΔT_g = 0.85 °C, ΔWBGT = 0.42 °C; autumn, ΔT_a = 0.55 °C, ΔT_g = 0.64 °C, ΔWBGT = 0.47 °C; winter, ΔT_a = 0.50 °C, ΔT_g = 0.58 °C, ΔWBGT = 0.50 °C, with the largest instantaneous differences occurring during summer early afternoon (13:00): ΔT_a = 1.37 °C, ΔT_g = 1.51 °C. Although these differences are more modest in magnitude than is sometimes reported for densely canopied sites, they are directionally consistent across all seasons and measurement days. The Kestrel 5400’s globe thermometer is also subject to the geometric limitations documented by Lamberti et al. [11], which may mean these shading differences slightly overstate the true radiative benefit of shading; this does not affect the directional conclusion but should be borne in mind when interpreting the specific magnitudes reported.

Diurnal patterns revealed seasonal peaks in parameter differences: in spring, the maximum differences in T_a and T_g occurred between 11:00 and 15:00, with WBGT peaking from 13:00 to 15:00; in summer, T_a and T_g peaked from 12:00 to 14:00, while WBGT peaks occurred at 9:00–10:00, 13:00–14:00, and 18:00–19:00; autumn peaks were observed at 12:00–14:00 and 19:00–21:00; in winter, ΔT_a was less stable, but T_g and WBGT peaked between 9:00 and 14:00. RH differences ranged from 2.0 to 3.39%, with maxima in early morning and evening, while wind speeds were generally higher in open areas than in shaded areas, with more pronounced differences at wind–corridor locations.

Based on these microclimate measurements, residents’ preferences for different sites were further analyzed. In spring, children and older adults preferred partially shaded or tree-shaded areas; in summer, activities for all age groups, particularly children and older adults, shifted to early or late hours, favoring shaded and well-ventilated zones; in autumn, clear light–thermal differences in the afternoon and evening, mainly related to solar exposure and shading, elicited heat-sensitive spatial choices; in winter, sun-exposed and wind-protected areas were preferred.

These observed patterns were translated into fine-grained design strategies, including increasing tree canopy coverage, pavilions, pergolas, and ventilated corridors, while preserving sunlit and wind-protected areas for winter, as summarized in

Table 7. Translating Microclimate Measurements into Population Activity Insights and Design Guidance.

Season	Microclimate Feature	Main Affected Population	Potential Behavioral Response	Design Implication
Spring	Shading begins to significantly reduce temperature and radiation around midday	Children, Older Adults	Prefer partially shaded or tree-shaded areas for activities	Maintain a combination of open and shaded spaces to create flexible multi-functional areas
Summer	Shading provides strongest cooling and radiation reduction; WBGT improvement most pronounced	All ages, especially children and older adults	Shift activity earlier or later; prioritize shaded and well-ventilated zones	Increase canopy coverage, pavilions, pergolas, and ventilation corridors to enhance shading
Autumn	Afternoon and evening still show noticeable light–thermal differences	Children, Adolescents, Adults	Spatial choice influenced by thermal conditions; increased evening stay	Strengthen west-facing sun control and optimize leisure spaces for evening use
Winter	Shading effect remains, but wind and solar exposure are more critical	Older Adults	Prefer wind-protected, sun-exposed, and stayable spaces	Retain partial shading while providing sunlit and wind-protected spaces for winter use.

[17,19,29,30].

3.3. Perception and Preference Analysis of Wind–Light–Thermal Conditions

Based on 1096 valid questionnaires, combined with field observations and microclimate measurements, this study conducted a comprehensive analysis of residents’ preferences and prioritization for wind, light, and thermal environments across spring, summer, autumn, and winter. Overall, residents’ environmental perception and preferences exhibited pronounced seasonal adaptation and age-related differences, reflecting the combined effects of microclimatic exposure, physiological regulation, and subjective environmental perception [1,12,13,17]. Wind and light conditions had the most significant influence on site selection, whereas thermal conditions were not the primary determinant in most cases. Figure 3 and Figure 4 illustrate the tendencies and priority distributions for wind, light, and thermal environments across seasons and age groups, while Figure 5 quantifies the proportion of respondents ranking each environmental factor as their top priority, together with the corresponding statistical tests.

In spring (Figure 3a and Figure 4a), most age groups showed a clear preference for transitional light zones, balancing sun exposure and shade, indicating a tendency to select spaces that provide both daylight access and shading during the moderate season. For wind, all age groups except older adults preferred light breeze zones, avoiding wind corridors or wind-exposed areas. Thermal conditions were generally rated as moderate, with children showing a relatively higher preference for warm conditions. Priority rankings further indicated that residents generally valued light and wind over thermal conditions, with children and older adults prioritizing light, while adolescents and some adults prioritized wind. Children assigned the highest importance to light (mean rank = 1.51), significantly higher than adolescents (mean rank = 1.92; post hoc Dunn’s test p < 0.001); in spring, only 12.5% of children ranked wind first, compared with 57.9% of adolescents (χ² = 21.51, p < 0.001). This suggests that during spring, when temperatures are moderate but wind and sunlight vary, site selection is jointly influenced by both daylight exposure and wind comfort.

In summer (Figure 3b and Figure 4b), preferences for light shifted significantly from transitional zones to shaded or semi-shaded areas across all age groups, especially for children, adults, and some older adults, reflecting the importance of shading under high temperatures and strong solar radiation. Wind remained a key factor, with most residents continuing to prefer light breeze zones and an increased proportion ranking wind as the top priority. Statistical analysis confirmed that 47.2% of all respondents placed wind as the first priority, higher than light (32.4%) and thermal conditions (20.4%), particularly pronounced among children, adults, and older adults. Children showed the most pronounced seasonal shift in wind priority: from 12.5% in spring to 58.8% in summer. Thermal comfort, while sometimes favoring cooler conditions for children and young adults, was not a primary factor, indicating that summer discomfort is mitigated mainly by seeking shade and breeze rather than direct thermal assessment.

In autumn (Figure 3c and Figure 4c), light preferences returned to transitional zones, and the proportion of respondents choosing such spaces increased across most age groups, reflecting decreased reliance on fully shaded spaces as solar radiation declined. Wind preference for light breeze zones remained dominant, indicating consistent demand for mild airflow. Thermal preferences were mostly moderate. Priority rankings indicated that wind and light were nearly equally important: 45.1% and 38.3% of respondents ranked wind and light first, respectively, compared with 16.7% for thermal conditions. Age differences persisted, with children and some older adults prioritizing light, while adolescents and some older adults prioritized wind. Age-group differences in wind-first proportion were not statistically significant in autumn (χ² = 2.49, p = 0.476). This suggests that in autumn, site selection shifts from summer-avoidance logic toward a balanced consideration of comfortable sunlight and wind.

Figure 5. Statistical analysis of environmental priority rankings, replacing the original Table 7. (a) Mean priority rank of wind, light, and thermal environments across age groups (Kruskal–Wallis test; p < 0.001). (b) Mean priority rank across four seasons. (c) Proportion of respondents ranking wind as top priority by age group and season (Pearson χ² test). (d) Seasonal pattern of wind, light, and thermal top-priority proportions for all respondents.

Figure 5. Statistical analysis of environmental priority rankings, replacing the original Table 7. (a) Mean priority rank of wind, light, and thermal environments across age groups (Kruskal–Wallis test; p < 0.001). (b) Mean priority rank across four seasons. (c) Proportion of respondents ranking wind as top priority by age group and season (Pearson χ² test). (d) Seasonal pattern of wind, light, and thermal top-priority proportions for all respondents.

In winter (Figure 3d and Figure 4d), residents’ light preferences shifted to sun-exposed areas, with most age groups showing strong preference for sunlight, except a few young adults. Shaded areas were generally avoided, indicating that solar access is the primary determinant under cold conditions. Wind preferences remained mostly for light breeze zones, but acceptance of no-wind areas increased, reflecting a need for wind protection. Thermal preferences were generally moderate; warm conditions were rarely accepted. Priority rankings reinforced this pattern: statistical analysis confirmed that 50.9% of total respondents ranked light as the top priority, higher than wind (32.2%) and thermal conditions (16.9%), particularly among children and older adults. Age-group differences in wind-first proportion were significant in winter (χ² = 12.49, p = 0.006): older adults showed the highest wind-first proportion (45.6%), while children showed the lowest (13.5%). This indicates that winter site selection is mainly driven by the combined need for sun exposure and wind protection, with thermal conditions acting indirectly rather than as an independent determinant.

Comparing Figure 3, Figure 4 and Figure 5, residents’ subjective environmental preferences were generally consistent with observed behavior. Field observations and environmental data analysis showed that residents adjusted their location and activity according to V_a and shading conditions, while fluctuations in T_a, RH, T_g, and WBGT within measured ranges did not significantly affect the number of participants or their distribution. Questionnaire-based priority rankings similarly showed that thermal conditions were rarely the top factor across seasons and age groups. Except for a few specific subgroups, the proportion of respondents ranking thermal conditions first was generally below 30%, with total respondents never exceeding 20.4% in any season. In contrast, wind and light consistently maintained high priority rankings, indicating that residents are more directly sensitive to wind, sunlight exposure, and shading conditions during routine outdoor activities. This descriptive pattern is statistically supported: Kruskal–Wallis tests confirmed significant age-group effects on light priority ranking (H = 56.78, df = 3, p < 0.001) and thermal priority ranking (H = 26.81, df = 3, p < 0.001), while wind priority ranking did not differ significantly across age groups (H = 4.83, df = 3, p = 0.185). Significant seasonal differences were found for wind ranking (H = 29.52, df = 3, p < 0.001) and light ranking (H = 27.11, df = 3, p < 0.001), but not for thermal ranking (H = 6.70, df = 3, p = 0.082), consistent with the interpretation that residents manage thermal discomfort primarily through behavioral adaptation rather than direct thermal assessment.

Overall preference patterns, summarized in Table 7 above and Figure 5, can be characterized as follows: in spring and autumn, transitional light-shade spaces are preferred; in summer, shaded and breezy areas are prioritized; in winter, sun-exposed and wind-protected areas are preferred. Age differences show that children and older adults are more sensitive to light, while adolescents and adults pay more attention to wind. Thermal conditions primarily influence site selection indirectly through interactions with light and wind, rather than serving as the primary determinant. Overall, wind and light can be considered key indicators for evaluating residential outdoor environmental quality, providing direct guidance for subsequent fine-grained analysis of activity patterns and site design strategies [1,12,27,30].

4. Discussion

Based on field observations, questionnaires, and microclimate measurements, this study systematically analyzed outdoor activity patterns and spatial selection behaviors of residents in typical Guangzhou residential neighborhoods across different seasons and age groups. Rather than restating the study aim and methodology here, this discussion focuses on interpreting the key findings in the context of existing literature and articulating the study’s theoretical contributions. Children and older adults engaged in sustained and widely distributed activities, whereas adolescents and adults exhibited highly concentrated activity patterns. Summer activities shifted earlier to avoid high temperatures, winter activities were concentrated at midday and afternoon, and spring and autumn showed three daily peaks in the morning, afternoon, and evening. These patterns indicate clear temporal adjustment strategies among different age groups, consistent with previous studies showing that sensitive populations adjust outdoor activity timing and spatial choices in response to heat exposure, solar radiation, and seasonal microclimatic variation [12,16,17,29]. A pattern that mirrors, at the scale of seasonal behavioral adjustment, what Xie et al. [6] observed at the scale of seconds: pedestrians’ thermal responses are not a fixed function of instantaneous conditions but depend on the immediate spatial and temporal context of exposure. Our finding that activity timing varies substantially between seasons even within a single age group can be read as the same underlying sensitivity to recent thermal context, operating over hours and days of seasonal exposure rather than the 20–35 s window Xie et al. identified at the pedestrian scale.

Activity type and spatial choice showed pronounced age differentiation. Children preferred playground areas and engaged in high-intensity activities; adolescents favored social or fitness spaces with seating opportunities; adults mainly performed low-intensity activities and participated in parent–child activities on non-working days; older adults preferred resting facilities and open plazas, while some also accompanied children to playgrounds or green spaces. Facility type, solar exposure, wind conditions, and site location collectively influenced spatial selection, and nighttime lighting significantly extended children’s activity periods, echoing previous evidence that open spaces, natural elements, and activity–related facilities can promote children’s outdoor activity, while age–friendly open spaces support older adults’ social interaction and daily use [13,16,29]. These findings suggest that fine-grained design should consider the multidimensional needs of different age groups in terms of functional spaces and microclimate conditions to enhance space utilization and diversity.

The regulatory effect of wind, light, and thermal environments on spatial selection was seasonally distinct. In spring and autumn, residents preferred transitional light-shade zones and light breeze areas; in summer, activities concentrated in shaded and well-ventilated zones; and in winter, activities favored sun-exposed and wind-protected areas. This study’s statistical results reinforce and sharpen this pattern: wind and light conditions were consistently prioritized over thermal conditions across all seasons and age groups, with thermal ranking showing no significant seasonal variation (H = 6.70, p = 0.082). This is consistent with evidence that behavioral and perceptual responses to outdoor thermal environments diverge from steady-state index predictions [19]: residents effectively convert thermal discomfort into spatial and temporal decisions—seeking shade (a light/radiation choice) and ventilated paths (a wind choice)—rather than responding to abstract thermal indices. The finding that thermal ranking is seasonally invariant, while wind and light rankings shift significantly, further supports this interpretation. This is consistent with studies showing that visual conditions, wind-related variables, and solar radiation can substantially affect outdoor thermal perception and environmental optimization across population groups [1,12,30]. Microclimate data further confirmed that tree shading significantly reduced T_a, T_g, and WBGT in summer, while sunlit and wind-protected areas remained the primary activity spaces in winter. Quantitatively, the present study found mean shading-related differences of ΔT_a = 0.50–0.95 °C, ΔT_g = 0.58–1.19 °C, and ΔWBGT = 0.30–0.50 °C, depending on the season—directionally consistent with, though more modest in magnitude than, prior work in hot and humid climates such as Yang and Lin [3] and Yang and Jian [28]. These results indicate that microclimate interventions can substantially improve outdoor comfort for different age groups across seasons, providing empirical support for environmentally responsive site design.

In terms of theoretical contribution, this study advances the literature in two principal ways. First, it provides an integrated behavioral–microclimatic evidence base linking age-specific activity patterns to environmental preferences at seasonal resolution, moving beyond the single-season or single-population focus of most existing studies. Second, the finding that thermal conditions are consistently de-prioritized across all seasons—while wind and light drive spatial selection—suggests that a thermal-comfort-index-centric design evaluation approach may underestimate the actual behavioral drivers in residential settings.

Based on these findings, this study proposes fine-grained design strategies for residential outdoor activity spaces, achieving a closed loop from data analysis to practical application. First, differentiated functional spaces should be provided for children, adolescents, adults, and older adults, such as playgrounds, pavilions, and fitness areas, with seasonal shading offered by trees or pergolas. Second, summer design should prioritize shaded zones and ventilated corridors to reduce thermal load, while winter spaces should provide sun exposure and wind protection to enhance comfort, which is in line with climate-responsive design studies emphasizing vegetation layout, semi-outdoor shading, airflow regulation, CFD-assisted microclimate optimization, and multi-objective green-space design [20,27,28,31]. Third, nighttime lighting should ensure safety and extend usable time for children and family activities. The ventilation corridor recommendation is grounded in residents’ expressed preference for light-breeze conditions and observed behavioral shifts toward ventilated spaces in summer, while the nighttime lighting recommendation is grounded in questionnaire data (25.3% of respondents) and field observations. By integrating microclimate improvements, wind-thermal preferences, and activity patterns, operational design guidelines can be established to support evidence-based fine-grained planning of residential outdoor spaces.

The generalizability of these findings beyond Guangzhou also warrants consideration. Guangzhou’s subtropical monsoon climate shapes both the severity of outdoor thermal stress and residents’ adaptive strategies; in temperate or arid climates, the relative priority of thermal, wind, and light factors would likely differ, although the core methodological framework—integrating questionnaire-based preference data with seasonal field measurements and age-stratified behavioral analysis—is expected to remain transferable.

Nevertheless, this study has several limitations. First, physiological monitoring was not included; activity preferences and thermal comfort were inferred solely from questionnaires and observations. Future studies could incorporate metrics such as skin temperature or heart rate to improve accuracy. Second, microclimate analysis was primarily based on six representative sites, and the applicability to other climate zones or building types requires further validation. Relatedly, the generalizability of these findings beyond Guangzhou is itself a limitation rather than a settled premise: Guangzhou’s subtropical monsoon climate shapes both the severity of outdoor thermal stress and residents’ adaptive strategies, so in temperate or arid climates the relative priority of thermal, wind, and light factors would likely differ from what is reported here, even though the underlying methodological framework—integrating questionnaire-based preference data with seasonal field measurements and age-stratified behavioral analysis—is expected to remain transferable. Third, we explicitly acknowledge the limitations of using globe thermometers to assess the outdoor radiant environment: globe thermometer measurements in open unshaded areas are subject to systematic Tmrt overestimation, as documented by Kántor and Unger [9], Tang et al. [10], and Lamberti et al. [11]; reported T_g and WBGT differences between shaded and open areas in the present study should therefore be interpreted as approximate rather than precise. Fourth, the ventilation corridor recommendation has not been verified through CFD simulation and should be treated as a design priority responsive to expressed resident need rather than an engineering-validated solution. Fifth, nighttime lighting design guidance is based on questionnaire responses and observational counts rather than measured illuminance data. Finally, future work could integrate measured data into dynamic comfort prediction, ENVI-met or CFD-based microclimate simulation, machine-learning-assisted evaluation, and multi-objective spatial optimization tools to provide more refined design guidance.

5. Conclusions

This study systematically analyzed outdoor activity patterns and spatial selection behaviors of residents in typical Guangzhou residential neighborhoods across different seasons and age groups, using questionnaires, field observations, and microclimate measurements. The regulatory effects of wind, light, and thermal environments on activity preferences were also examined. The main findings are summarized as follows:

• Activity patterns exhibit pronounced age differences and seasonal adjustments: Children and older adults engaged in sustained and widely distributed activities, whereas adolescents and adults displayed highly concentrated activity patterns. In summer, children’s peak activity window contracted by approximately 3 h relative to spring (to 08:00–10:00 and 16:00–19:00). Winter activities were concentrated at midday and afternoon, and spring and autumn showed three daily peaks in the morning, afternoon, and evening. These patterns indicate clear temporal adjustment strategies among different population groups.
• Activity types and spatial selection show significant differentiation: Children preferred playground areas and engaged in high-intensity activities; adolescents favored social or fitness spaces with seating opportunities; adults mainly performed low-intensity activities and participated in parent–child activities; older adults preferred resting facilities and open plazas. Facility type, solar exposure, wind conditions, and site location collectively influenced spatial choice, and nighttime lighting significantly extended children’s activity periods, consistent with questionnaire evidence that 25.3% of respondents reported that improved lighting would increase their evening outdoor time.
• Environmental priority rankings are significantly age- and season-dependent: In spring and autumn, residents preferred transitional light-shade zones and light-breeze areas; in summer, activities concentrated in shaded and well-ventilated areas; and in winter, sun-exposed and wind-protected areas were favored. Light priority ranking differed significantly across age groups (Kruskal–Wallis H = 56.78, p < 0.001), with children (mean rank = 1.51) and older adults (mean rank = 1.57) assigning it greater importance than adolescents (mean rank = 1.92). Wind priority differed significantly across seasons (H = 29.52, p < 0.001): 47.2% of respondents ranked wind first in summer, compared with 32.2% in winter. Thermal conditions were consistently deprioritized across all seasons and age groups (overall mean rank = 2.40), with no significant seasonal variation (p = 0.082). Tree shading significantly reduced Ta, Tg, and WBGT relative to open areas in every season (ΔT_a = 0.50–0.95 °C, ΔT_g = 0.58–1.19 °C, ΔWBGT = 0.30–0.50 °C), with the largest instantaneous gap during summer early afternoon (ΔT_g = 1.51 °C at 13:00), providing empirical support for improving outdoor comfort.
• Fine-grained site design enables evidence-based application: Based on activity patterns and microclimate analysis, this study proposes season- and population-specific design strategies, including functional zoning, shading and ventilation layouts, sun-exposed and wind-protected spaces, and optimized nighttime lighting. These strategies provide actionable guidance for fine-grained planning of residential outdoor activity spaces, contributing to improved comfort and health protection.

This study makes contributions at both theoretical and practical levels: by quantifying population differences, establishing a multidimensional environmental evaluation framework, and translating results into operational design strategies, it provides a systematic approach for residential outdoor space planning. Future research could further incorporate physiological monitoring data and microclimate simulation tools to develop more refined predictive models, enhancing the applicability of designs across different climate zones and building types.