1. Introduction
A growing concern with population inflation and urban expansion along with the heat island intensification, the lack of ventilation and decrease of greenspace in highly densified built up areas has resulted in deterioration of many climate-induced urban problems [
1,
2]. Still, the rapid urbanization is expected to emerge globally in the next a few decades which will lead to more and more people becoming urban inhabitants [
3], and inevitably cause the expanding demand for more living and working spaces. Meanwhile, the general trend of climate change is increasing pressures to the urban environments and posing environmental challenges to the urban planning for the outcome of public health and sustainable development. In addition, the awareness of the importance of urban bioclimatic conditions is growing along with the demand in the resilient and climate-responsive design for comfortable outdoor open spaces [
4,
5,
6,
7,
8,
9].
In contemporary society, urban inhabitants only spend less than 20% of their time out of buildings in some developed countries [
10]. It has long been recognized that comfortable outdoor public spaces meeting the expectations of occupants can attract people to spend more time in outdoor environments and substantially affect the likelihood of archiving sustainable urban development and enhancing the inhabitants’ quality of life [
11,
12,
13,
14,
15]. As the essential components in urban system, outdoor and semi-outdoor environments can provide open spaces for hosting various recreational, social and commercial activities [
16]. On the other hand, the outdoor green spaces in urban areas are facilitating to adapt to the increased heat stress and evolving into an important contribution to the energy efficiency of the surrounding buildings [
17,
18].
Since the last two decades, the research interest in assessing the outdoor thermal comfort in urban environments has drawn a great deal of worldwide attentions [
19]. From the literatures, rapid growth has come at the amounts of studies carried out in different geographical regions with distinct climates for developing and calibrating the model of comfort in outdoor urban spaces [
20,
21]. The rational indices, such as Physiological Equivalent Temperature (PET) [
22] and Universal Thermal Climate Index (UTCI) [
23,
24], have been widely applied for the comfort modeling and estimation of outdoor thermal conditions. However, solely relying on rational indexes is insufficient in providing full ranges of contextual and personal determinants on human comfort. The underpinning assumption of rational indices, which equates the heat balance between human body and surroundings with thermal comfort, was often violated and might give inconsistent results in empirical investigation. Great discrepancies were found between the calculated index value and the actual thermal sensations voted by the individual subjects [
7]. Besides a variety of microclimatic factors regarding thermal condition in outdoor environment, individuals’ perceptions of the environmental stimuli based on individual differences and features, and psychological and behavioral factors also contribute to the holistic assessment of subjective comfort in outdoor public spaces [
11,
25,
26,
27,
28].
As defined by ASHRAE, thermal comfort is that “state of mind in which human feels satisfied within the thermal environment” [
29]. The salient evolution of comfort model regarding psychological adaptation has been created referring to perception of and reaction to sensory information due to individuals’ experiences and expectations in a certain context [
20]. The active role of human agent has been revealed and recognized, which underpins the adaptive approach for thermal comfort modeling with consideration of human adaptations in terms of physiological, behavioral, and psychological dimensions [
30,
31,
32]. Further, the comprehensive conceptual model has been developed with expanded scope of influential factors, that indicates the ever-increasing importance of the personal physiological, psychological, and behavioral variables as well as non-thermal contextual factors [
25,
27,
33].
The outdoor comfort is rather difficult to be measured in physical or psychological quantities, thus, it is generally conceptualized as a latent construct in this study. The biometeorological index, as well as the individuals’ socio-demographical characteristics, long-established attitude towards urban public spaces, and momentary general environmental perception on the overall environment are needed to be synthesized to predict comfort in outdoor urban environment [
25,
33,
34,
35,
36]. By reviewing the conceptual models [
27,
33], we have included some influential factors for conceptualizing individuals’ momentary general environmental perception in urban public spaces, which may result in certain psychological effects on the holistic comfort. Moreover, the general environmental perception and the holistic comfort assessment are both hypothetically affected by individuals’ long-established attitudes towards urban public spaces, which reflects individuals’ experiences and socio-demographical backgrounds. To achieve a more comprehensive and robust model, an expanded set of factors are taken into account in our conceptual framework of comfort assessment. We address the new conceptual framework and conclude this study by proposing an approach to integrate individuals’ socio-demographical characteristics, long-established attitudes towards urban public spaces, momentary general environmental perceptions and rational thermal index into comfort modeling.
2. Conceptual Framework
The interplay between urban environmental condition and inhabitants’ quality of life makes the urban system as a whole. In turn, the investigations into outdoor comfort in urban public open spaces need combined and interdisciplinary approaches for gaining a holistic perspective. Individuals’ comfort perception is influenced by microclimatic and environmental stimuli to varying degrees based on psychological and behavioral adaptation, which is biased by their socio-demographical characteristics and living and working conditions. People with different demographic characteristics and socio-economic status are likely to experience different environments regarding certain behavioral patterns [
37]. Therefore, microclimatic, and environmental conditions need to be linked to individuals’ personal perceptions based on their different social and behavioral factors. The proposed conceptual framework emphasizes on the impacts from the long-established attitudes towards urban public spaces and momentary general environmental perceptions on the holistic comfort assessment in urban public spaces. The underlying assumption is that individuals’ attitudes and perceptions are determined by their social background and experience. As already presented in the previous conceptual model, both long-term and short-term components are effectual in comfort assessment [
27].
The measurements of comfort are not geographically or spatially invariable, which implies the importance of momentary perceptions regarding environmental features, and the contextual attributes regarding socio-cultural and climatological backgrounds. Many empirical evidences indicate the effects of the seasonal, geographical, and cultural differences on people’s thermal adaption in terms of physiological, psychological, and behavioral aspects [
38,
39,
40,
41]. During the outdoor activities, people gather multisensory experiences that inform their state of comfort in public realm [
42]. When applying an instrument of comfort assessment in different urban places, a key concern is to ensure if the measurement of the relevant constructs is spatially invariant, however, it is normally a failure because of the varying influences of individual’s environment perceptions [
43,
44]. The difference of climate responsive design strategies regarding urban geometry, planting vegetation, cool surface and water bodies in various urban open spaces has been addressed to improve the outdoor thermal comfort conditions [
5]. Apart from that, some spatial features of places or spots within a certain public space may have psychological effects on the occupants’ comfort assessment in different ways [
33,
34,
45]. From this point of view, we speculate the individual’s general perceptions on place-related differences may cause the non-independence of comfort assessment. From the literature, few studies systematically delve into this issue regarding environmental attitudes and perceptions.
The hypothetical conceptual framework is depicted in
Figure 1. In this diagram, the oral components denote latent variables regarding individuals’ attitudes towards urban public spaces, momentary general environmental perceptions and comfort assessment, which are constructed by indicators measured through the questionnaire. The exogenous variables regarding individuals’ socio-demographic characteristics are assumed to impact on these three latent variables. The thermal index calculated by measured meteorological variables hypothetically influences individuals’ outdoor comfort assessment as well. As shown in the diagram, the relationships among latent variables are proposed, which indicates that individuals’ momentary general environmental perceptions may influence comfort assessment. In addition, the long-established attitudes towards urban public spaces may impact on momentary general environmental perceptions and comfort assessment.
The diagram comes up with the detailed hypotheses regarding (I) the relationships between latent variables and manifest items, and (II) the relationships between exogenous factors and latent variables, which are listed as follow.
(I). hypothetical relationships between latent variables and manifest items:
Hypothesis 1a (H1a). Environmental attitude is measured by the answer to question “Do you agree that public green space is the most important infrastructure?”.
Hypothesis 1b (H1b). Environmental attitude is measured by the answer to question “Do you agree that public green space is conductive to spirit restoration and relaxation?”.
Hypothesis 1c (H1c). Environmental attitude is measured by the answer to question “Do you agree that open space is necessary in both residential neighborhoods and business districts?”.
Hypothesis 1c (H1d). Environmental attitude is measured by the answer to question “Do you agree that you prefer outdoor activities to indoor activities?”.
Hypothesis 1e (H1e). Environmental attitude is measured by the answer to question “Do you agree that people should spend more time for outdoor activities?”.
Hypothesis 1f (H1f). Environmental attitude is measured by the answer to question “Do you agree that recent weather is conductive to outdoor activities?”.
Hypothesis 1g (H1g). Environmental attitude is measured by the answer to question “Do you agree that more investments are needed to manage and maintain the urban public spaces?”.
Hypothesis 2a (H2a). Environmental perception is measured by the perception on green space in study area.
Hypothesis 2b (H2b). Environmental perception is measured by the perception on facilities in study area.
Hypothesis 2c (H2c). Environmental perception is measured by the perception on barrier-free design in study area.
Hypothesis 2d (H2d). Environmental perception is measured by the perception on hygienic condition of study area.
Hypothesis 2e (H2e). Environmental perception is measured by the perception on openness of study area.
Hypothesis 2f (H2f). Environmental perception is measured by sensation of noise in study area.
Hypothesis 2g (H2g). Environmental perception is measured by sensation of air quality in study area.
Hypothesis 3a (H3a). Comfort assessment is measured by thermal sensation in study area.
Hypothesis 3b (H3b). Comfort assessment is measured by the sensation of humidity in study area.
Hypothesis 3c (H3c). Comfort assessment is measured by the sensation of wind in study area.
Hypothesis 3d (H3d). Comfort assessment is measured by the sensation of radiation in study area.
Hypothesis 3e (H3e). Comfort assessment is measured by the sensation of sunlight in study area.
(II). hypothetical relationships between exogenous variables and latent variables:
Hypothesis 4 (H4). The effect of age on Environmental Attitude.
Hypothesis 5 (H5). The effect of gender on Environmental Attitude.
Hypothesis 6 (H6). The effect of education level on environmental attitude.
Hypothesis 7 (H7). The effect of income on environmental attitude.
Hypothesis 8 (H8). The effect of age on momentary environmental perception.
Hypothesis 9 (H9). The effect of gender on momentary environmental perception.
Hypothesis 10 (H10). The effect of education level on momentary environmental perception.
Hypothesis 11 (H11). The effect of income on momentary environmental perception.
Hypothesis 12 (H12). The effect of visiting frequency on momentary environmental perception.
Hypothesis 13 (H13). The effect of age on comfort assessment.
Hypothesis 14 (H14). The effect of gender on comfort assessment.
Hypothesis 15 (H15). The effect of education level on comfort assessment.
Hypothesis 16 (H16). The effect of income on comfort assessment.
Hypothesis 17 (H17). The effect of visiting frequency on comfort assessment.
Hypothesis 18 (H18). The effect of PET on comfort assessment.
Hypothesis 19 (H19). The effect of environmental attitude on momentary general environmental perception.
Hypothesis 20 (H20). The effect of environmental attitude on comfort assessment.
Hypothesis 21 (H21). The effect of momentary general environmental perception on comfort assessment.
5. Conclusions
This study presents a comprehensive conceptual model regarding the relationships between outdoor comfort and individuals’ long-established attitudes towards urban public spaces and momentary general environmental perceptions. A structural equation model was estimated using the data of 372 subjects surveyed in two public spaces in Changsha city. Most of the hypothetical relationships proposed in the conceptual framework are verified. Unlike previous studies, individual’s holistic comfort assessment is conceptualized as a latent variable, which is unmeasurable and can only be measured by the indicators regarding thermal sensation, and sensations of radiation and sunlight during the surveys in study areas. The sensations of wind and humidity measured in the surveys are not significantly correlated with comfort assessments in the local context as presented in the results. Nevertheless, as a latent construct, individual’s holistic comfort assessment is expected to be measured by specific sensations in different context of geographical regions.
The results of SEM estimate provide quantitative evidence, which indicates physical thermal exposure condition is the strong effect on individual’s comfort assessment. Meanwhile, the important role of person-related variables in outdoor comfort modeling has been revealed. The mechanism of comfort perception involves the long-established attitude towards urban public spaces and the momentary general environmental perception in accordance with the previous conceptual model of comfort perception proposed by Lenzholzer and de Vrijs (2019) [
27]. Individual’s comfort in urban public spaces is not only based on the current state when the comfort perception is recorded but also the attitudes established in the outdoor experience and socio-demographical factors.
Unlike previous empirical investigations focusing on the momentary influential factors only, this study emphasizes on the importance of individual’s socio-demographical characteristics and long-term established psychological factors in outdoor comfort modeling. More empirical evidence related to the respondents with various socio-demographical backgrounds in different geographical regions are expected to be carried out in the coming future.