- Email: [email protected]
Climate and behaviour in a Nordic city
Landscape and Urban Planning 82 (2007) 72–84
Climate and behaviour in a Nordic city Ingeg¨ard Eliasson a,∗ , Igor Knez b , Ulla Westerberg b , Sofia Thorsson a,b , Fredrik Lindberg a a
Urban Climate Group, Physical Geography, Department of Earth Sciences, G¨oteborg University, Box 460, SE-405 30 G¨oteborg, Sweden b Department of Technology and Built Environment, University of G¨ avle, Box 88, SE-80176 G¨avle, Sweden Received 29 June 2006; received in revised form 25 January 2007; accepted 31 January 2007 Available online 13 March 2007
Abstract Four urban public spaces, representing various designs and microclimates, were investigated in Gothenburg, Sweden, in order to estimate how weather and microclimate affect people in urban outdoor environments. The research strategy was both multidisciplinary and interdisciplinary and included scientists from three disciplines: architecture, climatology and psychology. The project is based on common case studies carried out during four seasons, including measurements of meteorological variables, interviews and observations of human activity at each place. Multiple regression analysis of meteorological and behavioural data showed that air temperature, wind speed and clearness index (cloud cover) have a significant influence on people’s assessments of the weather, place perceptions and place-related attendance. The results support the arguments in favour of employing climate sensitive planning in future urban design and planning projects, as the physical component of a place can be designed to influence the site-specific microclimate and consequently people’s place-related attendance, perceptions and emotions. © 2007 Elsevier B.V. All rights reserved. Keywords: Public spaces; Urban climate; Climate planning; Environmental psychology; Environmental design
1. Introduction
1.1. Climate and urban design
Scientists from a wide range of different disciplines including architecture, climatology, engineering and psychology have long been interested in how weather and climate affect people in the urban outdoor environment. Several factors have been shown to influence people’s perceptions and use of the outdoor environment, among them the design and function of the space, as well as the physiological and psychological parameters involved in human reactions to the physical environment. To date, most of the research has been carried out within the individual disciplines. As a result, the different factors have been identified, but knowledge of their individual and combined influence is still lacking, since an integrated research approach is necessary for such analyses.
Two disciplines, architecture and urban design and urban climatology, dominate the published literature on how buildings and the urban environment affect climate (Mills, 1999). A key objective within architecture and urban design is the creation of a ‘comfortable’ living environment. Research on this topic often has a bioclimatic focus and an empirical and inferential approach and the results are normally presented as guidelines and realworld examples. In contrast, research in urban climatology, a special field within meteorology and climatology, focuses on measurements and the modelling of physical processes in order to interpret the changes in atmospheric properties that give rise to the “urban effect”. With some exceptions, research within urban climatology is not carried out for the purposes of design and the results obtained are often theoretical and not readily interpretable from a design perspective (Mills, 1999, 2006; Eliasson, 2000). Bioclimatic urban design is pointed out as a potential subject for research in which the combined skills of the climatologist and the designer can be beneficially employed. One of the classical
∗
Corresponding author. Tel.: +46 31 7732832; fax: +46 31 7731986. E-mail addresses: [email protected] (I. Eliasson), [email protected] (I. Knez), [email protected] (U. Westerberg), [email protected] (S. Thorsson), [email protected] (F. Lindberg). 0169-2046/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.landurbplan.2007.01.020
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
lines of research within human biometeorology is the development of comfort indices that model and predict the thermal interaction between the human body and its surrounding environment (e.g. H¨oppe and Seidl, 1991). Over the years, outdoor thermal indices have been criticised mainly due to their inability to provide realistic assessment under transient exposure and to include psychological factors. As shown by Nikolopoulou and Steemers (2003), only approximately 50% of the variance between objective and subjective comfort evaluations could be explained by the physical and physiological conditions. They suggest other factors that could influence the tolerance interval for thermal comfort, such as experience, expectations, sense of control, the “naturalness” of the environment and the need for stimulation.
73
1.3. The purpose of this paper This paper describes an investigation of urban public places in relation to micrometeorological variations and human perceptions of climate. The study combines meteorological and behavioural data in an analysis of the impact of three weather variables (clearness index (CI), air temperature (Ta ) and wind (w)) on participants’ perceptions of current weather and on their behavioural, aesthetical and emotional assessments of four urban public spaces. The main goal of the investigation was to test the hypothesis that the three weather variables have a significant influence on peoples’ weather assessments and placerelated perceptions, emotions and attendance. 2. Methods
1.2. Climate and human psychology 2.1. An integrative research approach Research on how emotion, cognition and activity influence the tolerance range for climate comfort is comparatively rare. The relation between functional use and microclimatic conditions has been confirmed by several studies (e.g. Gehl, 1971; Westerberg, 1994; Nikolopoulou et al., 2001; Zacharias et al., 2001; Thorsson et al., 2004, 2006) which show that comfortable weather conditions, i.e., high temperature and access to sunlight increases the number of people present in an urban space. Studies also show that both too cold and too warm conditions have a negative influence on the emotional state, which in turn tends to trigger aggressive behaviour (e.g. Cohn, 1993; Simister and Cooper, 2005). Emotional and cognitive research suggests that emotional states can influence cognitive processes (e.g. Blaney, 1986; Kuiken, 1991). If climate is a moderator of emotional state, then it is likely that it also affects other aspects of the environmental experience, such as the visual aesthetics (e.g. Gifford, 1980; Knez and Thorsson, 2006). There also appears to be a link between thermal comfort and some psychological aspects of the environmental experience (see Knez and Thorsson, 2006). The concept of space, comprising physical and spatial connotations, has traditionally been used in geographical and architectural discourse. It does not include the psychological and social aspects of spatial experiences and has therefore been redefined in environmental psychology by the notion of place (e.g. Graumann, 2002). Several authors have sketched similar accounts of the theory of a place (e.g. Canter, 1977) comprising three key components: physical (form and space), functional (activities) and psychological (meanings people assign to a place) aspects. Canter (1997) has further developed his earlier model into four “facets” of place: functional differentiation, place objectives, scale of interaction and aspects of design. Place objectives extend the previous psychological (individual) aspects by including both social and cultural components, while the scale of interaction addresses the environmental aspects. Yet, as pointed out by Knez (2005), an insufficiency in these theoretical accounts is the omission of climate, which influences individual, social, economic (Parker, 1995), and criminal behaviour (Rotton and Cohn, 2002) and memories of, and meanings we attribute to places (Knez, 2006).
The present study is part of the “Urban Climate Spaces” project, involving scientists from the fields of climatology, psychology and architecture. The project has an integrative research approach with a common goal: its aim is to traverse disciplinary boundaries in order to develop integrated knowledge and theory, i.e., to conduct interdisciplinary work (Tress et al., 2004) for the purposes of analysing the complex relational links between climate and human behaviour and its implications for sustainable urban design (Knez, 2005, 2006; Knez and Thorsson, 2006, 2007; Lindberg, 2005; Thorsson et al., 2006). 2.2. Case studies Case studies were conducted in the city of Gothenburg, which is located at latitude 57◦ N on the Swedish west coast. Four urban public spaces with different design and varying microclimates ranging from an exposed waterfront plaza and large open square to a park with shading trees and a small sheltered courtyard were included in the study (Fig. 1). Micrometeorological measurements, observations and structured interviews were conducted simultaneously during four case studies (October 2003, January, April and June 2004). Each study period included five days over a period of two weeks. The aim was to find 5 days in each season with different weather, with respect to air temperature, cloud cover and wind speed. Rainy days were excluded. In total, 20 days of measurements, observations and interviews were conducted in Gothenburg. The case studies were performed between 11 a.m. and 3 p.m. Solar radiation and temperature normally reach their daily maxima during this period, and the places under study are frequently used by people. 2.3. Micrometeorological measurements The air temperature and relative humidity (Rotronic YA-100), globe temperature (AMR Pt100 PK24) as well as incoming short and long wave radiation (Kipp and Zonen CM3 and CG1) were measured at a height of 1.1 m above the ground, corresponding to the average height of the centre of gravity for adults (Mayer and H¨oppe, 1987). Wind speed and direction (Gill Ultransonic)
74
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 1. Photographs of the four study areas and the characteristics of the microclimate at each place.
were measured at a height of 2 m and later recalculated to represent a measurement height of 1.1 m using Sverdrup’s power law. The instruments were mounted on two carriers. One carrier was always situated at the square, while the other was rotated daily between the three other study areas. The clearness index (CI) was calculated from radiation measurements (Kipp and Zonen, CNR 1) at a reference station located above roof level (32 m above ground). The clearness index is defined as the ratio between a measured and a theoretical maximum incoming solar radiation for a specific time and location on the earth’s surface. High values of CI (e.g. >0.75) represents clear sky conditions, while lower values represent more cloudy conditions. 2.4. Observations and structured interviews Observations of human activity and behaviour were made every 20 min, at the same time as the meteorological measurements and interviews. Physical activity, i.e., the number of people lying, sitting, standing and walking as well as visible qualifications such as eating, talking and reading were observed. The structured interview contained eight main questions about demographic variables, clothing, general and specific questions about current weather, behaviour, feelings and attitudes related to the site (Thorsson et al., 2006). People were randomly approached and the questionnaire took about five minutes to complete. A total of 1379 people participated in the study (560 at the square, 351 at the courtyard, 266 in the park and 202 at the waterfront plaza). There were similar numbers of women and men visiting each place. About 80% of the participants were between 21 and 65 years of age. Three main measures of dependency, that is, three questions from the questionnaire were analysed in this study. These concerned the estimations of current weather and the behavioural and perceptual dimensions of a place. The first question was related to the current weather, i.e., “What is your perception of the weather today?”. Participants were asked to respond to three 5-point scales ranging from 1 to
5: (1) calm–windy, (2) cold–warm and (3) good–bad for outdoor activity (see Knez and Thorsson, 2006; Thorsson et al., 2006). The second question was related to the place, i.e., “What is your perception of the place at this moment?”. Participants were asked to respond according to four 5-point scales ranging from 1 to 5: (1) ugly–beautiful, (2) unpleasant–pleasant, (3) windy–calm and (4) cold–warm (see Knez and Thorsson, 2006; Thorsson et al., 2006). The third question was related to the emotional states of (de)activation and (dis)pleasure, i.e., “How do you feel in this place at this moment?” Participants had to respond according to four 5-point scales ranging from 1 to 5: (1) elated–bored, (2) glad–gloomy, (3) calm–nervous, and (4) active–passive. These scales were derived from the Knez and Hygge (2001) measure of the current effect. In connection to this question participants were also asked to estimate their thermal comfort by responding according to a 9-point scale ranging from very cold to very hot, with the score 5 indicating “comfortable” (Matzarakis and Mayer, 1996). 2.5. Multiple regression analyses Multiple regression analyses were performed to investigate the influence of the three independent physical variables (clearness index, CI; air temperature, Ta ; wind speed, w) on the participants’ evaluations of current weather and their behavioural and perceptual estimations of each place (the dependent variables i.e. the three above-mentioned questions from the questionnaire). This statistical technique may be viewed as a descriptive and/or inferential instrument by which the (linear) influence of the three independent variables on each dependent variableis evaluated collectively and separately. In other words, the amount of variance in a dependent variable (criterion) that could be attributed to the three independent variables (predictors) jointly and separately was analysed. The analysis is based on 1-min mean data for wind speed, the air temperature measured at each site and 5-min mean data for the clearness index;
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
the index was calculated from radiation measurements which were taken at the roof of the reference station. The meteorological data were synchronised to the start time of each interview. It must be noted that 5-min mean data for wind speed were also analysed, however there was no difference between these results and those obtained using the 1-min mean data. In order to determine whether the independent variables correlated, i.e., to check for multicollinearity, the variance inflation factor (VIF) was calculated. Multicollinearity would result in a greater error variance in the multiple regression model and is considered to be severe when VIF is greater than 10 (e.g. Pfaffenberger and Patterson, 1987). However, the results of the multiple regression analyses described below showed VIFs between 1.1 and 1.9, indicating a very low (and for the multiple regression models, satisfactory) level of intercorrelation between the independent variables. 3. Results The results are reported and discussed with a focus on each of the four urban public spaces. Results from the multiple regression analyses are presented in Tables 2–13 for each place and dependent variable (question from the questionnaire). Due to the huge amount of statistical data involved, only significant results are discussed. 3.1. Meteorological statistics The city of Gothenburg is located on the outskirts of the marine west-coast climate region and has a mean air temperature of 16.3 ◦ C in July and −0.5 ◦ C in January (Fig. 2). The upper diagram in Fig. 2 also shows the range around the mean daytime maximum air temperature value, measured at the reference station during the four case study periods in January, April, June and October. The mean daytime maximum air temperature
75
during the case studies in winter (January), summer (June) and autumn (October) were low in comparison to the 30-year daily maximum value for Gothenburg. In fact, the coldest event in the month of October since 1901 (−8.5 ◦ C) occurred during the autumn measurements. On the other hand, the mean daytime maximum air temperature during the spring measurements taken in April were higher than the 30-year mean value (Fig. 2). The lowest diagram in Fig. 2 shows the variation of the Clearness Index during the four case studies. The autumn case study covers the widest range of CI, followed by the winter case study in January. In comparison, the spring and summer case studies included more days with clear skies (Fig. 2). The air temperature variation within each study area for each season is showed in Fig. 3a. It is notable that the air temperature during the summer case study is only slightly higher than during the spring case studies which is an effect of the unusually warm April in 2004 as shown previously in Fig. 2. From Fig. 3a, it is evident that the air temperature variation between the different study areas is very small for all seasons. However, measurements at the square show somewhat higher maximum temperatures, especially during winter and spring case studies (Fig. 3a). The variable weather during the autumn case study (Fig. 2) is reflected in the range of the air temperature at the four places (Fig. 3a). The variation in wind speed within each place for each season is shown in Fig. 3b. The mean wind speed was much higher at the waterfront in comparison to the other three places, especially in the summer, due to one windy day. On this day – but also at most other days – the wind speed at the waterfront plaza was almost as high as at the reference station. This was probably due to large fetch distances from the western to northern to north-easterly directions at the waterfront plaza. The largest variation in wind speed between the four places occurred during the winter case study (Fig. 3b). The highest mean wind speed was recorded in the park; however, the highest maximum wind speed value was recorded at the waterfront plaza. During the other three seasons, the wind speed variation was relatively small between the square, park and courtyard. The wind speed values shown in Fig. 3b are generally below 4 m/s and are often less than 2 m/s, which may appear to be low values. However, it must be remembered that the wind speed values in Fig. 3b were measured at a height of 2.0 m above ground. 3.2. Results from study area 1: the square
Fig. 2. Upper diagram: monthly mean air temperature and mean daily minimum and maximum temperature in Gothenburg (1961–1990). The average daily maximum air (Ta ) during the measurement is also included. Lower diagram: the clearness index (CI) during the field measurements.
3.2.1. Question 1: “What is your perception of the weather today?” The results of the multiple regression analysis showed that the number of people (attendance) in the square increased with increasing clearness index (β = 0.18) and at higher air temperatures (β = 0.51), with R2 = 0.33 (see Table 1). Technically, β shows the direction and strength of the slope between Y and X and R2 indicates how much of the variance in Y is explained by X. Thus the results indicate that the clearness index and air temperature accounted for 33% of the variance in attendance at the square. The results also show that the wind speed and the air temperature had a significant influence
76
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 3. Upper diagram (a) shows box plots of air temperature (Ta ) and lower diagram (b) shows box plots of wind speed (w) from the four study areas in Gothenburg during the field measurements (N = number of days). The bottom and top edges of the boxes represent the 25th and 75th percentiles, respectively. The “whiskers” extend to the 5th and 95th percentiles. The “+” sign indicates the minimum and maximum values.
Table 1 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the total attendance and perceived “current weather” at the square Dependent variable
R2
Total attendance -CI -Ta
.33
Calm–windy -w -Ta
.17
Cold–warm -w -Ta
.39
Good–bad for outdoor activity -CI -w -Ta
.15
Beta
df
MS
F
3,174
10,929.8
28.56
.000 .009 .000
3,559
62.11
39.09
.000 .000 .000
3,559
116.66
120.62
.000 .000 .000
3,559
35.62
31.31
.000 .000 .001 .000
.18 .51 .33 .15 −.19 .65 −.23 .15 −.21
p
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
77
Table 2 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the perceived “place right now” at the square Dependent variable
R2
Ugly–beautiful -w -Ta
.06
Unpleasant–pleasant -CI -Ta
.07
Windy–calm -CI -w -Ta
.11
Cold–warm -CI -w -Ta
.14
Beta
df
MS
F
p
3,559
12.95
10.84
.000 .000 .002
3,559
13.01
13.06
.000 .030 .000
3,559
34.34
21.68
.000 .001 .000 .023
3,557
45.05
29.55
.000 .000 .016 .000
−.18 .15 .10 .22 .15 −.22 −.10 .19 −.11 .27
on the participants’ assessments of the current weather on the “calm–windy” and “cold–warm” dimensions. More precisely, the current weather was estimated as warmer at lower wind speeds and higher air temperatures, and 39% of the variance in this type of weather estimation was due to the wind and air temperature. The current weather was estimated as calmer when both the wind speed and air temperature decreased, indicating a relatively low R2 value of 0.17. In addition, participants estimated weather as being better for outdoor activity at times with clear skies, high air temperature and decreasing wind speed (R2 = 0.15). 3.2.2. Question 2: “What is your perception of the place at this moment?” Results showed that the participants perceived the square to be more beautiful at times with low wind speed and high air temperatures, indicating a low R2 value of 0.06 (see Table 2). Regarding the R2 values, it must be noted that results related to the aesthetical evaluations of the places and how participants felt at these places are generally low (see Tables 3 and 4). This
means that although the weather parameters are shown to significantly influence the dependent variables, they do not explain much of the variance in these measurements. Such a result is quite common in behavioural sciences due to the: (a) measurement errors in subjective assessment and (b) the large body of uncontrolled variables that may also have influenced the participants’ aesthetical assessments and why they felt as they did when visiting a place. As shown in Table 2, the square was also rated as being more pleasant when the clearness index and air temperature increased. The results also show that the participants estimated this place as being more calm and warm at times with low wind speeds, clear skies and low air temperature. 3.2.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 3, the participants felt more positive (that is, more elated, happier, calmer) in the square at higher air temperatures. In addition, they felt more elated on occasions with clear skies and more thermally comfortable at lower wind speeds.
Table 3 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the perceived “feelings right now” at the square Dependent variable
R2
Elated–bored -CI -Ta
.04
Glad–gloomy -Ta
.04
Calm–nervous -Ta
.03
Comfort scale -w -Ta
.11
Beta
df
MS
F
3,558
6.98
8.3
.000 .005 .004
3,558
6.23
7.29
.000 .000
3,558
5.3
6.02
.000 .001
3,270
11.17
10.75
.000 .004 .000
−.13 −.14 −.17 −.16 −.17 .27
p
78
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Table 4 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the total attendance and perceived “current weather” at the courtyard Dependent variable
R2
Total attendance -CI -w -Ta
.49
Beta
df
MS
F
p
365
2883.19
20.17
.000 .043 .001 .000
3,350
37.3
28.21
.000 .000 .000
3,350
59.99
67.27
.000 .000 .008 .000
3,350
15.79
17.89
.000 .000 .000
.20 −.32 .63
Calm–windy -w -Ta
.20
Cold–warm -CI -w -Ta
.37
Good–bad for outdoor activity -CI -Ta
.13
.30 .25 −.16 −.12 .65 −.22 −.26
3.3. Results from study area 2: the courtyard 3.3.1. Question 1: “What is your perception of the weather today?” The results showed that the total attendance at the courtyard increased when the wind ceased and the air temperature and solar radiation increased, accounting for almost 50% of the variance (see Table 4). The results also showed that the participants perceived the current weather to be warmer and better for outdoor activity when the air temperature rose. The current weather, on the other hand, was perceived as being colder at higher wind speeds, lower air temperature and with clearer skies, accounting for 37% of the variance. Participants assessed the weather as being more windy at higher wind speeds. However, the analysis also showed that “windy weather” can be related to an increase in air temperature, which is probably a local, place-related phenomenon. It was also observed that clear sky weather was perceived to be important for outdoor activity. 3.3.2. Question 2: “What is your perception of the place at this moment?” The results showed that the participants rated the courtyard as being more beautiful at higher wind speeds and lower air
temperatures (see Table 5). The courtyard was also assessed as being a more pleasant place when the wind speed and clearness index increased. The participants assessed the courtyard as being a calmer place when the wind speed decreased, but lower air temperatures also turned out to have a significant influence on the participants’ assessment of this place’s “windiness”. 3.3.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 6, the participants felt happier and calmer at lower wind speeds at the courtyard. The results also showed that they felt more active at the courtyard during conditions with clear skies and high temperatures. 3.4. Results from study area 3: “the park” 3.4.1. Question 1: “What is your perception of the weather today?” The results showed that the total attendance in the park increased at times when the weather was warm (see Table 7). The results also showed that the participants assessed the current weather as being more windy when wind speed increased. In addition, they perceived the current weather to be warmer as a result of higher air temperature.
Table 5 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the perceived “place right now” at the courtyard Dependent variable
R2
Ugly–beautiful -w -Ta
.04
Unpleasant–pleasant -CI -w
.03
Windy–calm -w -Ta
.07
Beta
df
MS
F
p
3,350
3.22
5.19
.002 .000 .044
3,350
1.52
4.07
.007 .013 .013
3,350
11.69
8.11
.000 .003 .005
.20 −.12 .14 .14 −.16 −.16
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
79
Table 6 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the perceived “feelings right now” at the courtyard Dependent variable
R2
Glad–gloomy -w
.02
Calm–nervous -w
.02
Active–passive -CI -Ta
.08
Beta
df
MS
F
p
3,350
1.58
2.61
.051 .017
3,350
1.76
2.79
.040 .005
3,350
17.77
10.35
.000 .001 .001
−.13 −.16 .17 .19
Table 7 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the total attendance and perceived “current weather” in the park Dependent variable
R2
Total attendance -w -Ta
.36 .27
Calm–windy -w
.28
Beta
df
MS
F
p
3,56
517.21
9.71
.000 .020 .000
3,265
47.67
33.95
.000 .000
3,265
98.98
167.35
.000 .026 .000
3,265
10.66
12.57
.000 .005 .014 .020
.57 .53
Cold–warm -w -Ta
.66 .09 .82
Good–bad for outdoor activity -CI -w -Ta
.13 −.18 .16 −.14
The participants also assessed the current weather as being “good for outdoor activity” at times with clear skies, high air temperatures and low wind speeds.
the park as a “calm” and “warm” place. The results also showed that the park was evaluated as a “warm” place at higher air temperatures.
3.4.2. Question 2: “What is your perception of the place at this moment?” The results showed that the participants perceived the park to be a more beautiful place on occasions with clear skies and low air temperatures (see Table 8). Low wind speeds and clear skies were shown to have a significant influence on the perceptions of
3.4.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 9, the participants felt more bored and gloomy at times when the sky was cloudy in the park. In addition, the participants felt more passive on occasions with high air temperature.
Table 8 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the perceived “place right now” in the park Dependent variable
R2
Ugly–beautiful -CI -Ta
.05
Windy–calm -CI -w
.26
Cold–warm -CI -w -Ta
.24
Beta
df
MS
F
3,265
4.04
4.85
.003 .051 .021
3,263
42.69
31.10
.000 .049 .000
3,264
35.88
27.76
.000 .000 .000 .033
.13 −.16 .12 −.45 .26 −.27 .12
p
80
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Table 9 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the perceived “feelings right now” in the park Dependent variable
R2
Elated–bored -CI
.03
Glad–gloomy -w
.05
Active–passive -Ta
.19
Comfort scale -w
.05
Beta
df
MS
F
p
3,263
2.51
2.78
.041 .007
3,265
2.48
4.32
.005 .008
3,264
33.41
19.88
.000 .000
3,159
2.98
−.18 −.18 .43 .17
2.68 .042
.049
Table 10 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the total attendance and perceived “current weather” at the waterfront plaza Dependent variable
R2
Total attendance -w -Ta
.22
Beta
df
MS
F
3,45
75.67
3.9
.015 .028 .004
3,201
45.51
36.73
.000 .000 .000
3,201
16.71
19.80
.000 .002 .000
3,201
14.80
15.89
.000 .000 .000
−.43 .56
Calm–windy -CI -w
.36
Cold–warm -w -Ta
.23
Good–bad for outdoor activity -CI -w
.19
−.23 .58 −.26 .60 −.41 .30
3.5. Results from study area 4: the waterfront plaza 3.5.1. Question 1: “What is your perception of the weather today?” The results showed that the number of participants at the waterfront plaza increased at times with lower wind speeds and higher temperatures (see Table 10). At this place, the participants perceived the current weather to be more windy, colder and worse for outdoor activity when the wind speed increased. Clearer skies, however, led the participants to assess the current weather as being calmer and better for outdoor activity. The weather was perceived to be warmer when the air temperature rose.
p
3.5.2. Question 2: “What is your perception of the place at this moment?” As shown in Table 11, the participants assessed the waterfront plaza as being more beautiful at higher wind speeds and lower air temperature. The waterfront plaza was also assessed as a “warmer” and “calmer” place on occasions with clear skies and when the wind speed increased. 3.5.3. Question 3: “How do you feel in this place at this moment?” The results showed that only the active–passive behavioural dimension was significantly influenced by the three weather variables (see Table 12). More precisely, the participants at the
Table 11 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the perceived “place right now” at the waterfront plaza Dependent variable
R2
Ugly–beautiful -w -Ta
.04
Windy–calm -CI -w
.21
Cold–warm -CI
.05
Beta
df
MS
F
3,201
2.82
2.98
.033 .004 .022
3,201
23.06
17.04
.000 .005 .000
3,201
3.81
3.23
.024 .008
.27 −.21 .19 −.43 .19
p
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
81
Table 12 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the perceived “feelings right now” at the waterfront plaza Dependent variable
R2
Active–passive -CI -w -Ta
.10
Beta
df
MS
F
p
3, 201
13.49
7.64
.000 .014 .030 .000
.18 −.20 .38
4.1. Weather parameters and the functional component
Fig. 4. The complex interrelation between weather and microclimate and the three components of place, adapted from Canter (1977).
waterfront plaza felt more active at high wind speeds, lower air temperatures and cloudy skies. 4. Discussion In general, the results of this study confirm the hypothesis that the air temperature, wind speed and clearness index (i.e., cloud cover) have a significant influence on the participants’ weather assessments and place-related perceptions, emotions and attendance. The results clearly show that weather and microclimate have a significant influence on two (functional and psychological) of three components constituting a place (Fig. 4).
Regarding the functional component of a place, measured as the participants’ attendance, the air temperature was shown to have a similarly significant influence in all places. Fig. 5 gives a schematic illustration of this impact (reported in Tables 2, 5, 8 and 11). It is clear from the diagrams in Fig. 5 that when the air temperature rose, the number of visitors increased in all four places. The statistical analysis not only showed that these results were significant, it also showed that 50% of the variance in attendance could be explained by weather parameters (CI, Ta , w), where Ta explains up to 63% of the variance (Table 4). Another general result for all four places was that the participants classified clear sky weather and weather with weak winds as being better for outdoor activity. A schematic illustration of this effect, based on results presented in Tables 2, 5, 8 and 11, is presented in Fig. 6. These results are in line with previous studies performed in Scandinavia (Carlestam, 1968; Gehl and Yencken, 1996; Thorsson, 2003; Thorsson et al., 2004) and other parts of Europe (Nikolopoulou and Steemers, 2003; Nikolopoulou and Lykoudis, 2006), which demonstrated a strong influence of weather on the use of outdoor urban spaces. The weather and climate in Scandinavia are characterised as both variable and changeable. It is also well-known that in these countries, the dark winter and the following short summer have fostered a special relationship between the inhabitants and the sun (Gehl and Yencken, 1996). If it is possible to enjoy the sun for only a short period, the urge to do so is tremendous. Results from a study that was conducted by the authors of this article in Matsudo, a satellite city of Tokyo, Japan, using the same methods as those presented above, showed a comparably low influence of weather on the use of different outdoor spaces (Thorsson et al., 2006). This difference between Sweden and Japan is probably due to both cultural and climatological differences (Knez
Fig. 5. Total attendance as a function of Ta , CI and w (square, courtyard, park and waterfront plaza). The diagrams are a schematic illustration of the statistics reported in Tables 2, 5, 8 and 11.
82
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 6. Plots of good–bad settings for outdoor activity as a function of Ta , CI and w (square, courtyard, park and waterfront plaza). The diagrams are a schematic illustration of the statistics reported in Tables 2, 5, 8 and 11.
and Thorsson, 2007). Large differences in comfort conditions between different cities in Europe were also reported from the RUROS project (Nikolopoulou and Lykoudis, 2006). 4.2. Weather parameters and the psychological component It is known that architects, urban planners and climatologists have long argued that climate-related issues are important for city life. But how do the people living in the city, i.e., “the public”, relate to climatic elements, and are weather and climate of any importance to an individual (Eliasson, 2000)? Based on the results presented in this study, this question must be answered in the affirmative. It is very clear from the study that the people visiting the four places do care about, and are influenced by, the weather and climate. Thus, weather parameters are important—not only did they influence the number of people who visited the spaces and whether they assessed the weather as being good or bad for outdoor activity, but they also influenced how people assessed the places and how being in these places made them feel, i.e., the psychological component of a place. Concerning the psychologically-related results, and as pointed out in Section 3, it must be noted that only tentative links between participants’ behaviour, weather and place were indicated, due to the low R2 values. This type of result is related to measurement errors in subjective assessment and to the large body of uncontrolled variables that may also have had an impact on participants’ psychological responses. An interesting result is that the open, exposed waterfront plaza was estimated to be more beautiful as a result of high wind speed and low air temperatures, as shown in Fig. 7. Participants also felt more active when there was higher wind speed in the waterfront plaza (Table 2). In contrast, participants visiting the open square estimated it to be more beautiful as a result of lower wind speeds and
Fig. 7. Ugly–beautiful as a function of w at the square (left diagram) and waterfront plaza (right diagram). The diagrams are a schematic illustration of the statistics reported in Tables 3, 6, 9 and 12.
higher air temperatures. A tentative explanation is that wind at the waterfront renders a positive aesthetic and symbolic value. This place gains life through the natural elements reflected in the water and the activities at the adjacent marina become more interesting to look at on a windy day. The attraction of the square, and to some extent even the park, are other people. The typical park or square activities are merely reduced by the wind. This corresponds with the theory put forward by Westerberg (1994), who illustrated that gustiness caused by unsuitable building geometry was disliked much more than the blustery winds on hilltops or natural maritime settings, where even “bad” weather would become an attraction. Nikolopoulou and Steemers (2003) arrive at the same conclusions: in urban spaces with high amounts of natural characteristics such as parks, the tolerance to widespread changes in the physical environment is high, provided they are produced naturally. Several other studies show that the experience of nature in the city is a source of positive feelings (e.g., Chiesura, 2004). The results also show that feelings of pleasantness increase with higher clearness index. Clear skies also increased the subjective weather-related assessments of warmth and calm in three of the four urban spaces (square, park and waterfront plaza, Table 1). 4.3. Weather parameters and the physical component Weather and climate influence the physical component of a place. However, as illustrated in Fig. 4, the relationships also work in the opposite direction. A building’s shape, orientation, material, colour, etc., influence radiation, temperature, wind and other parameters, to produce a site-specific microclimate. Urban designers/planners are thus able to create urban environments that take advantage of the positive effects of the existing weather and climate. Climate design has always been a natural part of local building tradition. However, when designing the modern city, the urban designer/planner has to consider many different conflicting aspects. Thus climate-related issues have a relatively small impact on the planning process (Eliasson, 2000). We hope, however, that studies such as this one will serve to bolster the arguments in favour of using climate-sensitive planning in future designs of cities. What are these arguments? Usage and activity are often used as measures of successful urban spaces. For example, Carmona et al. (2003, p. 99) states that “successful public places are characterised by the presence of people”. The present study shows that in order to create successful public places, weather parameters cannot
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
be neglected, since they account for over 50% of the variance in place-related attendance. Viewed together with results illustrating the influence of weather parameters on perceptions and emotions, it is clear that climate-sensitive planning can be an important tool in the drive to increase sustainability in the urban environment. Attractive urban spaces influence the social life in the city and will also indirectly influence the local economy and transportation infrastructure. In this perspective, climatesensitive planning may influence all of the three dimensions of sustainability, i.e., environmental, social and economic qualities. More research on the human response to urban climate is however needed before this theory can be proved. The compact city versus an urban sprawl is one of the key issues facing future urban sustainable development, both in Sweden and abroad. It is important to remember that climate-sensitive planning of urban public places is important, regardless of which future direction is chosen. In practice, climate-sensitive planning is a matter of choosing between alternatives. It is difficult to apply one general rule to all types of urban spaces. Regardless of whether it is a new construction or a re-design, a starting point for climate-sensitive planning is to take advantage of the seasonal variations of the microclimate. Urban spaces can be “climatically” attractive all year round. One way is to design a differentiated microclimate within the place. As each urban space is unique, it is important to first identify environmental characteristics such as openness, shadow pattern and wind conditions within the place and then simulate the proposed changes. Several interesting projects have recently developed GIS-based models as tools for analysing environmental diversity (e.g. Ratti and Richens, 2004; Steemers, 2006). 5. Conclusion Weather parameters (clearness index, air temperature and wind) have a significant influence on participants’ weather assessments and place-related perceptions, emotions, attendance. It is thus clear that solar radiation, air temperature and wind are vital aspects of the functional and psychological components of a place. The results support the arguments in favour of employing climate-sensitive planning in future urban design and planning projects, as the physical component of a place can be designed to influence the site-specific microclimate and consequently people’s place-related attendance, perceptions and emotions. Acknowledgements The project is financially supported by Formas, the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning in their key action area Urban Public Spaces. References Blaney, P.H., 1986. Affect and memory: a review. Psychol. Bull. 99, 229–246. Canter, D., 1977. The Psychology of Place. The Architectural Press Ltd., London.
83
Canter, D., 1997. The facets of place. In: Moore, G.T., Marams, R.W. (Eds.), Advances in Environment, Behavior, and Design. Plenum press, New York, pp. 109–147. Carlestam, G., 1968. Studier av utomhusaktiviteter med automatisk kamera. Rapport R16:68, Byggforskningsr˚adet. Carmona, M., Heat, T., Oc, T., Tiedsell, S., 2003. Public Places—Urban Spaces: The Dimensions of Urban Design. Archtectural Press, Elsevier. Chiesura, A., 2004. The role of urban parks for the sustainable city. Landsc. Urban Planning 68, 129–138. Cohn, E.G., 1993. The prediction of police calls for service: the influence of weather and temporal variables on rape and domestic violence. J. Environ. Psychol. 13, 71–83. Eliasson, I., 2000. The use of climate knowledge in urban planning. Landsc. Urban Planning 48, 31–44. Gehl, J., 1971. Livet mellem husene. K¨openhamn. Gehl, J., Yencken, D., 1996. Byens rum—byens liv. Arkitektens forlag. Kunstakademiets forlag. Gifford, R., 1980. Environmental dispositions and the evaluation of architectural interiors. J. Res. Pers. 14, 386–399. Graumann, C.F., 2002. The phenomenological approach to people-environment studies. In: Bechtel, R.B., Churchman, A. (Eds.), Handbook of Environmental Psychology. John Wiley & Sons, Inc., New York, pp. 95–113. H¨oppe, P., Seidl, H.A.J., 1991. Problems in assessment of the bioclimate for vacationists at the seaside. Int. J. Biometeorol 35, 107–110. Knez, I., 2005. Attachment and identity as related to a place and its perceived climate. J. Environ. Psychol. 25, 207–218. Knez, I., 2006. Autobiographical memories for places. Memory 14, 359–377. Knez, I., Hygge, S., 2001. The circumplex structure of affect: a Swedish version. Scand. J. Psychol. 42, 389–398. Knez, I., Thorsson, S., 2006. Influence of culture and environmental attitude on thermal, emotional and perceptual evaluations of a square. Int. J. Biometeorol 50 (5), 258–268. Knez, I., Thorsson, S., 2007. Thermal, emotional and perceptual evaluations of a park: cross-cultural and environmental attitude comparisons. Environ. Behav., submitted for publication. Kuiken, D., 1991. Mood and Memory: Theory, Research and Applications. Sage, London. Lindberg, F., 2005. Towards the use of local governmental 3-D data within urban climatology studies. Mapping Image Sci. 2005 (2), 4–9. Matzarakis, A., Mayer, H., 1996. Another kind of environmental stress. WHO News 18, 7–10. Mayer, H., H¨oppe, P., 1987. Thermal comfort of man in different urban environments. Theor. Appl. Clim. 38, 43–49. Mills, G., 1999. Urban Climatology and Urban Design. ICUC, International Conference on Urban Climatology in Sydney, Australia, 8–12 November. Extended abstracts. Mills, G., 2006. Progress toward sustainable settlements: a role for urban climatology. Theor. Appl. Climatol. 84, 69–76. Nikolopoulou, M., Baker, N., Steemers, K., 2001. Thermal comfort in outdoor urban spaces: understanding the human parameter. Solar Energy 70 (3), 227–235. Nikolopoulou, M., Steemers, K., 2003. Thermal comfort and psychological adaptation as a guide for designing urban spaces. Energy Buildings 35, 95–101. Nikolopoulou, M., Lykoudis, S., 2006. Thermal comfort in outdoor urban spaces: analysis across different European countries. Building Environ. 41, 1455–1470. Parker, P.M., 1995. Climatic Effects on Individual, Social and Economic Behavior. Greenwood Press, Westport, CT. Pfaffenberger, R.C., Patterson, J.H., 1987. Statistical Methods. Irwin, Homewood, IL. Ratti, C., Richens, P., 2004. Raster analysis of urban form. Environ. Planning B Planning Des. 31 (2), 297–306. Rotton, J., Cohn, E., 2002. Climate, weather and crime. In: Bechtel, R.B., Churchman, A. (Eds.), Handbook of Environmental Psychology. John Wiley & Sons, Inc., New York, pp. 481–498. Simister, J., Cooper, C., 2005. Thermal stress in the U.S.A. effects on violence and on employee behaviour. Stress Health 21, 3–15.
84
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Steemers, K., 2006. Human comfort in urban spaces. In: The 6th International Conference on Urban Climate (ICUC6). Thorsson, S., 2003. Climate, air-quality and thermal comfort in the urban environment. Doctoral Thesis A87. G¨oteborg University, Sweden. Thorsson, S., Lindqvist, M., Lindqvist, S., 2004. Thermal bioclimatic conditions and patterns of behaviour in an urban park in Sweden. Int. J. Biometeorol. 48, 149–156. Thorsson, S., Honjo, T., Lindberg, F., Eliasson, I., Eun-Mi, L., 2006. Thermal comfort and outdoor activity in Japanese urban public spaces. Environ. Behav., in press. Tress, G., Tress, B., Fry, G., 2004. Clarifying integrative research concepts in landscape ecology. Landsc. Ecol. 20, 479–493. Westerberg, U., 1994. Climatic planning—physics or symbolism. Architecture Behav. 19, 49–72. Zacharias, J., Stathopoulos, T., Wu, H., 2001. Microclimate and downtown open space activity. Environ. Behav. 33, 296–315. Ingeg¨ard Eliasson Born on 1961, presented her PhD thesis in 1993 and is since 2000 associate professor in physical geography at G¨oteborg University. Her current research focus on studies of wind and turbulence field and energy balance in urban street canyons, human biometeorology and thermal comfort, studies of urban climate and air pollution in African cities and integration of climate knowledge in urban planning.
Igor Knez Born on 1959, has, since 1992, a PhD in psychology from the University of Uppsala, Sweden. Since 1998, he is associate professor in psychology at Centre for Built Environment, University of G¨avle, Sweden. His current environmental psychology research focuses on place-related identity and memory and place-related psychological assessment. Ulla Westerberg Born on 1946, received her PhD in 1993 on climatic design in architecture. She has been the project leader of urban climate and housing surveys and has cooperated in projects on energy efficiency and environmental assessment of buildings at Centre for Built Environment, G¨avle University. Sofia Thorsson Born on 1972, received her PhD in 2003 in physical geography. She went on post-doc at Chiba University, Japan and University of G¨avle, in 2004 and 2005. Since 2005 she is post-doctoral Fellow at G¨oteborg University. Research focus is on urban climate, human biometeorology, thermal comfort and air pollution in urban areas. Fredrik Lindberg Born on 1974, received his MA at G¨oteborg University in 2002 and is currently a PhD-candidate in physical geography at G¨oteborg University, Sweden. His main expertises are urban climatology, human biometeorology and geographical information science.
Climate and behaviour in a Nordic city Ingeg¨ard Eliasson a,∗ , Igor Knez b , Ulla Westerberg b , Sofia Thorsson a,b , Fredrik Lindberg a a
Urban Climate Group, Physical Geography, Department of Earth Sciences, G¨oteborg University, Box 460, SE-405 30 G¨oteborg, Sweden b Department of Technology and Built Environment, University of G¨ avle, Box 88, SE-80176 G¨avle, Sweden Received 29 June 2006; received in revised form 25 January 2007; accepted 31 January 2007 Available online 13 March 2007
Abstract Four urban public spaces, representing various designs and microclimates, were investigated in Gothenburg, Sweden, in order to estimate how weather and microclimate affect people in urban outdoor environments. The research strategy was both multidisciplinary and interdisciplinary and included scientists from three disciplines: architecture, climatology and psychology. The project is based on common case studies carried out during four seasons, including measurements of meteorological variables, interviews and observations of human activity at each place. Multiple regression analysis of meteorological and behavioural data showed that air temperature, wind speed and clearness index (cloud cover) have a significant influence on people’s assessments of the weather, place perceptions and place-related attendance. The results support the arguments in favour of employing climate sensitive planning in future urban design and planning projects, as the physical component of a place can be designed to influence the site-specific microclimate and consequently people’s place-related attendance, perceptions and emotions. © 2007 Elsevier B.V. All rights reserved. Keywords: Public spaces; Urban climate; Climate planning; Environmental psychology; Environmental design
1. Introduction
1.1. Climate and urban design
Scientists from a wide range of different disciplines including architecture, climatology, engineering and psychology have long been interested in how weather and climate affect people in the urban outdoor environment. Several factors have been shown to influence people’s perceptions and use of the outdoor environment, among them the design and function of the space, as well as the physiological and psychological parameters involved in human reactions to the physical environment. To date, most of the research has been carried out within the individual disciplines. As a result, the different factors have been identified, but knowledge of their individual and combined influence is still lacking, since an integrated research approach is necessary for such analyses.
Two disciplines, architecture and urban design and urban climatology, dominate the published literature on how buildings and the urban environment affect climate (Mills, 1999). A key objective within architecture and urban design is the creation of a ‘comfortable’ living environment. Research on this topic often has a bioclimatic focus and an empirical and inferential approach and the results are normally presented as guidelines and realworld examples. In contrast, research in urban climatology, a special field within meteorology and climatology, focuses on measurements and the modelling of physical processes in order to interpret the changes in atmospheric properties that give rise to the “urban effect”. With some exceptions, research within urban climatology is not carried out for the purposes of design and the results obtained are often theoretical and not readily interpretable from a design perspective (Mills, 1999, 2006; Eliasson, 2000). Bioclimatic urban design is pointed out as a potential subject for research in which the combined skills of the climatologist and the designer can be beneficially employed. One of the classical
∗
Corresponding author. Tel.: +46 31 7732832; fax: +46 31 7731986. E-mail addresses: [email protected] (I. Eliasson), [email protected] (I. Knez), [email protected] (U. Westerberg), [email protected] (S. Thorsson), [email protected] (F. Lindberg). 0169-2046/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.landurbplan.2007.01.020
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
lines of research within human biometeorology is the development of comfort indices that model and predict the thermal interaction between the human body and its surrounding environment (e.g. H¨oppe and Seidl, 1991). Over the years, outdoor thermal indices have been criticised mainly due to their inability to provide realistic assessment under transient exposure and to include psychological factors. As shown by Nikolopoulou and Steemers (2003), only approximately 50% of the variance between objective and subjective comfort evaluations could be explained by the physical and physiological conditions. They suggest other factors that could influence the tolerance interval for thermal comfort, such as experience, expectations, sense of control, the “naturalness” of the environment and the need for stimulation.
73
1.3. The purpose of this paper This paper describes an investigation of urban public places in relation to micrometeorological variations and human perceptions of climate. The study combines meteorological and behavioural data in an analysis of the impact of three weather variables (clearness index (CI), air temperature (Ta ) and wind (w)) on participants’ perceptions of current weather and on their behavioural, aesthetical and emotional assessments of four urban public spaces. The main goal of the investigation was to test the hypothesis that the three weather variables have a significant influence on peoples’ weather assessments and placerelated perceptions, emotions and attendance. 2. Methods
1.2. Climate and human psychology 2.1. An integrative research approach Research on how emotion, cognition and activity influence the tolerance range for climate comfort is comparatively rare. The relation between functional use and microclimatic conditions has been confirmed by several studies (e.g. Gehl, 1971; Westerberg, 1994; Nikolopoulou et al., 2001; Zacharias et al., 2001; Thorsson et al., 2004, 2006) which show that comfortable weather conditions, i.e., high temperature and access to sunlight increases the number of people present in an urban space. Studies also show that both too cold and too warm conditions have a negative influence on the emotional state, which in turn tends to trigger aggressive behaviour (e.g. Cohn, 1993; Simister and Cooper, 2005). Emotional and cognitive research suggests that emotional states can influence cognitive processes (e.g. Blaney, 1986; Kuiken, 1991). If climate is a moderator of emotional state, then it is likely that it also affects other aspects of the environmental experience, such as the visual aesthetics (e.g. Gifford, 1980; Knez and Thorsson, 2006). There also appears to be a link between thermal comfort and some psychological aspects of the environmental experience (see Knez and Thorsson, 2006). The concept of space, comprising physical and spatial connotations, has traditionally been used in geographical and architectural discourse. It does not include the psychological and social aspects of spatial experiences and has therefore been redefined in environmental psychology by the notion of place (e.g. Graumann, 2002). Several authors have sketched similar accounts of the theory of a place (e.g. Canter, 1977) comprising three key components: physical (form and space), functional (activities) and psychological (meanings people assign to a place) aspects. Canter (1997) has further developed his earlier model into four “facets” of place: functional differentiation, place objectives, scale of interaction and aspects of design. Place objectives extend the previous psychological (individual) aspects by including both social and cultural components, while the scale of interaction addresses the environmental aspects. Yet, as pointed out by Knez (2005), an insufficiency in these theoretical accounts is the omission of climate, which influences individual, social, economic (Parker, 1995), and criminal behaviour (Rotton and Cohn, 2002) and memories of, and meanings we attribute to places (Knez, 2006).
The present study is part of the “Urban Climate Spaces” project, involving scientists from the fields of climatology, psychology and architecture. The project has an integrative research approach with a common goal: its aim is to traverse disciplinary boundaries in order to develop integrated knowledge and theory, i.e., to conduct interdisciplinary work (Tress et al., 2004) for the purposes of analysing the complex relational links between climate and human behaviour and its implications for sustainable urban design (Knez, 2005, 2006; Knez and Thorsson, 2006, 2007; Lindberg, 2005; Thorsson et al., 2006). 2.2. Case studies Case studies were conducted in the city of Gothenburg, which is located at latitude 57◦ N on the Swedish west coast. Four urban public spaces with different design and varying microclimates ranging from an exposed waterfront plaza and large open square to a park with shading trees and a small sheltered courtyard were included in the study (Fig. 1). Micrometeorological measurements, observations and structured interviews were conducted simultaneously during four case studies (October 2003, January, April and June 2004). Each study period included five days over a period of two weeks. The aim was to find 5 days in each season with different weather, with respect to air temperature, cloud cover and wind speed. Rainy days were excluded. In total, 20 days of measurements, observations and interviews were conducted in Gothenburg. The case studies were performed between 11 a.m. and 3 p.m. Solar radiation and temperature normally reach their daily maxima during this period, and the places under study are frequently used by people. 2.3. Micrometeorological measurements The air temperature and relative humidity (Rotronic YA-100), globe temperature (AMR Pt100 PK24) as well as incoming short and long wave radiation (Kipp and Zonen CM3 and CG1) were measured at a height of 1.1 m above the ground, corresponding to the average height of the centre of gravity for adults (Mayer and H¨oppe, 1987). Wind speed and direction (Gill Ultransonic)
74
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 1. Photographs of the four study areas and the characteristics of the microclimate at each place.
were measured at a height of 2 m and later recalculated to represent a measurement height of 1.1 m using Sverdrup’s power law. The instruments were mounted on two carriers. One carrier was always situated at the square, while the other was rotated daily between the three other study areas. The clearness index (CI) was calculated from radiation measurements (Kipp and Zonen, CNR 1) at a reference station located above roof level (32 m above ground). The clearness index is defined as the ratio between a measured and a theoretical maximum incoming solar radiation for a specific time and location on the earth’s surface. High values of CI (e.g. >0.75) represents clear sky conditions, while lower values represent more cloudy conditions. 2.4. Observations and structured interviews Observations of human activity and behaviour were made every 20 min, at the same time as the meteorological measurements and interviews. Physical activity, i.e., the number of people lying, sitting, standing and walking as well as visible qualifications such as eating, talking and reading were observed. The structured interview contained eight main questions about demographic variables, clothing, general and specific questions about current weather, behaviour, feelings and attitudes related to the site (Thorsson et al., 2006). People were randomly approached and the questionnaire took about five minutes to complete. A total of 1379 people participated in the study (560 at the square, 351 at the courtyard, 266 in the park and 202 at the waterfront plaza). There were similar numbers of women and men visiting each place. About 80% of the participants were between 21 and 65 years of age. Three main measures of dependency, that is, three questions from the questionnaire were analysed in this study. These concerned the estimations of current weather and the behavioural and perceptual dimensions of a place. The first question was related to the current weather, i.e., “What is your perception of the weather today?”. Participants were asked to respond to three 5-point scales ranging from 1 to
5: (1) calm–windy, (2) cold–warm and (3) good–bad for outdoor activity (see Knez and Thorsson, 2006; Thorsson et al., 2006). The second question was related to the place, i.e., “What is your perception of the place at this moment?”. Participants were asked to respond according to four 5-point scales ranging from 1 to 5: (1) ugly–beautiful, (2) unpleasant–pleasant, (3) windy–calm and (4) cold–warm (see Knez and Thorsson, 2006; Thorsson et al., 2006). The third question was related to the emotional states of (de)activation and (dis)pleasure, i.e., “How do you feel in this place at this moment?” Participants had to respond according to four 5-point scales ranging from 1 to 5: (1) elated–bored, (2) glad–gloomy, (3) calm–nervous, and (4) active–passive. These scales were derived from the Knez and Hygge (2001) measure of the current effect. In connection to this question participants were also asked to estimate their thermal comfort by responding according to a 9-point scale ranging from very cold to very hot, with the score 5 indicating “comfortable” (Matzarakis and Mayer, 1996). 2.5. Multiple regression analyses Multiple regression analyses were performed to investigate the influence of the three independent physical variables (clearness index, CI; air temperature, Ta ; wind speed, w) on the participants’ evaluations of current weather and their behavioural and perceptual estimations of each place (the dependent variables i.e. the three above-mentioned questions from the questionnaire). This statistical technique may be viewed as a descriptive and/or inferential instrument by which the (linear) influence of the three independent variables on each dependent variableis evaluated collectively and separately. In other words, the amount of variance in a dependent variable (criterion) that could be attributed to the three independent variables (predictors) jointly and separately was analysed. The analysis is based on 1-min mean data for wind speed, the air temperature measured at each site and 5-min mean data for the clearness index;
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
the index was calculated from radiation measurements which were taken at the roof of the reference station. The meteorological data were synchronised to the start time of each interview. It must be noted that 5-min mean data for wind speed were also analysed, however there was no difference between these results and those obtained using the 1-min mean data. In order to determine whether the independent variables correlated, i.e., to check for multicollinearity, the variance inflation factor (VIF) was calculated. Multicollinearity would result in a greater error variance in the multiple regression model and is considered to be severe when VIF is greater than 10 (e.g. Pfaffenberger and Patterson, 1987). However, the results of the multiple regression analyses described below showed VIFs between 1.1 and 1.9, indicating a very low (and for the multiple regression models, satisfactory) level of intercorrelation between the independent variables. 3. Results The results are reported and discussed with a focus on each of the four urban public spaces. Results from the multiple regression analyses are presented in Tables 2–13 for each place and dependent variable (question from the questionnaire). Due to the huge amount of statistical data involved, only significant results are discussed. 3.1. Meteorological statistics The city of Gothenburg is located on the outskirts of the marine west-coast climate region and has a mean air temperature of 16.3 ◦ C in July and −0.5 ◦ C in January (Fig. 2). The upper diagram in Fig. 2 also shows the range around the mean daytime maximum air temperature value, measured at the reference station during the four case study periods in January, April, June and October. The mean daytime maximum air temperature
75
during the case studies in winter (January), summer (June) and autumn (October) were low in comparison to the 30-year daily maximum value for Gothenburg. In fact, the coldest event in the month of October since 1901 (−8.5 ◦ C) occurred during the autumn measurements. On the other hand, the mean daytime maximum air temperature during the spring measurements taken in April were higher than the 30-year mean value (Fig. 2). The lowest diagram in Fig. 2 shows the variation of the Clearness Index during the four case studies. The autumn case study covers the widest range of CI, followed by the winter case study in January. In comparison, the spring and summer case studies included more days with clear skies (Fig. 2). The air temperature variation within each study area for each season is showed in Fig. 3a. It is notable that the air temperature during the summer case study is only slightly higher than during the spring case studies which is an effect of the unusually warm April in 2004 as shown previously in Fig. 2. From Fig. 3a, it is evident that the air temperature variation between the different study areas is very small for all seasons. However, measurements at the square show somewhat higher maximum temperatures, especially during winter and spring case studies (Fig. 3a). The variable weather during the autumn case study (Fig. 2) is reflected in the range of the air temperature at the four places (Fig. 3a). The variation in wind speed within each place for each season is shown in Fig. 3b. The mean wind speed was much higher at the waterfront in comparison to the other three places, especially in the summer, due to one windy day. On this day – but also at most other days – the wind speed at the waterfront plaza was almost as high as at the reference station. This was probably due to large fetch distances from the western to northern to north-easterly directions at the waterfront plaza. The largest variation in wind speed between the four places occurred during the winter case study (Fig. 3b). The highest mean wind speed was recorded in the park; however, the highest maximum wind speed value was recorded at the waterfront plaza. During the other three seasons, the wind speed variation was relatively small between the square, park and courtyard. The wind speed values shown in Fig. 3b are generally below 4 m/s and are often less than 2 m/s, which may appear to be low values. However, it must be remembered that the wind speed values in Fig. 3b were measured at a height of 2.0 m above ground. 3.2. Results from study area 1: the square
Fig. 2. Upper diagram: monthly mean air temperature and mean daily minimum and maximum temperature in Gothenburg (1961–1990). The average daily maximum air (Ta ) during the measurement is also included. Lower diagram: the clearness index (CI) during the field measurements.
3.2.1. Question 1: “What is your perception of the weather today?” The results of the multiple regression analysis showed that the number of people (attendance) in the square increased with increasing clearness index (β = 0.18) and at higher air temperatures (β = 0.51), with R2 = 0.33 (see Table 1). Technically, β shows the direction and strength of the slope between Y and X and R2 indicates how much of the variance in Y is explained by X. Thus the results indicate that the clearness index and air temperature accounted for 33% of the variance in attendance at the square. The results also show that the wind speed and the air temperature had a significant influence
76
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 3. Upper diagram (a) shows box plots of air temperature (Ta ) and lower diagram (b) shows box plots of wind speed (w) from the four study areas in Gothenburg during the field measurements (N = number of days). The bottom and top edges of the boxes represent the 25th and 75th percentiles, respectively. The “whiskers” extend to the 5th and 95th percentiles. The “+” sign indicates the minimum and maximum values.
Table 1 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the total attendance and perceived “current weather” at the square Dependent variable
R2
Total attendance -CI -Ta
.33
Calm–windy -w -Ta
.17
Cold–warm -w -Ta
.39
Good–bad for outdoor activity -CI -w -Ta
.15
Beta
df
MS
F
3,174
10,929.8
28.56
.000 .009 .000
3,559
62.11
39.09
.000 .000 .000
3,559
116.66
120.62
.000 .000 .000
3,559
35.62
31.31
.000 .000 .001 .000
.18 .51 .33 .15 −.19 .65 −.23 .15 −.21
p
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
77
Table 2 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the perceived “place right now” at the square Dependent variable
R2
Ugly–beautiful -w -Ta
.06
Unpleasant–pleasant -CI -Ta
.07
Windy–calm -CI -w -Ta
.11
Cold–warm -CI -w -Ta
.14
Beta
df
MS
F
p
3,559
12.95
10.84
.000 .000 .002
3,559
13.01
13.06
.000 .030 .000
3,559
34.34
21.68
.000 .001 .000 .023
3,557
45.05
29.55
.000 .000 .016 .000
−.18 .15 .10 .22 .15 −.22 −.10 .19 −.11 .27
on the participants’ assessments of the current weather on the “calm–windy” and “cold–warm” dimensions. More precisely, the current weather was estimated as warmer at lower wind speeds and higher air temperatures, and 39% of the variance in this type of weather estimation was due to the wind and air temperature. The current weather was estimated as calmer when both the wind speed and air temperature decreased, indicating a relatively low R2 value of 0.17. In addition, participants estimated weather as being better for outdoor activity at times with clear skies, high air temperature and decreasing wind speed (R2 = 0.15). 3.2.2. Question 2: “What is your perception of the place at this moment?” Results showed that the participants perceived the square to be more beautiful at times with low wind speed and high air temperatures, indicating a low R2 value of 0.06 (see Table 2). Regarding the R2 values, it must be noted that results related to the aesthetical evaluations of the places and how participants felt at these places are generally low (see Tables 3 and 4). This
means that although the weather parameters are shown to significantly influence the dependent variables, they do not explain much of the variance in these measurements. Such a result is quite common in behavioural sciences due to the: (a) measurement errors in subjective assessment and (b) the large body of uncontrolled variables that may also have influenced the participants’ aesthetical assessments and why they felt as they did when visiting a place. As shown in Table 2, the square was also rated as being more pleasant when the clearness index and air temperature increased. The results also show that the participants estimated this place as being more calm and warm at times with low wind speeds, clear skies and low air temperature. 3.2.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 3, the participants felt more positive (that is, more elated, happier, calmer) in the square at higher air temperatures. In addition, they felt more elated on occasions with clear skies and more thermally comfortable at lower wind speeds.
Table 3 Results from multiple regression analyses (N = 560) on the influence of the clearness index (CI = 0.1–0.91), wind speed (w = 0.35–8.2 ms−1 ) and air temperature (Ta = −7.1 to 20.98 ◦ C) on the perceived “feelings right now” at the square Dependent variable
R2
Elated–bored -CI -Ta
.04
Glad–gloomy -Ta
.04
Calm–nervous -Ta
.03
Comfort scale -w -Ta
.11
Beta
df
MS
F
3,558
6.98
8.3
.000 .005 .004
3,558
6.23
7.29
.000 .000
3,558
5.3
6.02
.000 .001
3,270
11.17
10.75
.000 .004 .000
−.13 −.14 −.17 −.16 −.17 .27
p
78
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Table 4 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the total attendance and perceived “current weather” at the courtyard Dependent variable
R2
Total attendance -CI -w -Ta
.49
Beta
df
MS
F
p
365
2883.19
20.17
.000 .043 .001 .000
3,350
37.3
28.21
.000 .000 .000
3,350
59.99
67.27
.000 .000 .008 .000
3,350
15.79
17.89
.000 .000 .000
.20 −.32 .63
Calm–windy -w -Ta
.20
Cold–warm -CI -w -Ta
.37
Good–bad for outdoor activity -CI -Ta
.13
.30 .25 −.16 −.12 .65 −.22 −.26
3.3. Results from study area 2: the courtyard 3.3.1. Question 1: “What is your perception of the weather today?” The results showed that the total attendance at the courtyard increased when the wind ceased and the air temperature and solar radiation increased, accounting for almost 50% of the variance (see Table 4). The results also showed that the participants perceived the current weather to be warmer and better for outdoor activity when the air temperature rose. The current weather, on the other hand, was perceived as being colder at higher wind speeds, lower air temperature and with clearer skies, accounting for 37% of the variance. Participants assessed the weather as being more windy at higher wind speeds. However, the analysis also showed that “windy weather” can be related to an increase in air temperature, which is probably a local, place-related phenomenon. It was also observed that clear sky weather was perceived to be important for outdoor activity. 3.3.2. Question 2: “What is your perception of the place at this moment?” The results showed that the participants rated the courtyard as being more beautiful at higher wind speeds and lower air
temperatures (see Table 5). The courtyard was also assessed as being a more pleasant place when the wind speed and clearness index increased. The participants assessed the courtyard as being a calmer place when the wind speed decreased, but lower air temperatures also turned out to have a significant influence on the participants’ assessment of this place’s “windiness”. 3.3.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 6, the participants felt happier and calmer at lower wind speeds at the courtyard. The results also showed that they felt more active at the courtyard during conditions with clear skies and high temperatures. 3.4. Results from study area 3: “the park” 3.4.1. Question 1: “What is your perception of the weather today?” The results showed that the total attendance in the park increased at times when the weather was warm (see Table 7). The results also showed that the participants assessed the current weather as being more windy when wind speed increased. In addition, they perceived the current weather to be warmer as a result of higher air temperature.
Table 5 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the perceived “place right now” at the courtyard Dependent variable
R2
Ugly–beautiful -w -Ta
.04
Unpleasant–pleasant -CI -w
.03
Windy–calm -w -Ta
.07
Beta
df
MS
F
p
3,350
3.22
5.19
.002 .000 .044
3,350
1.52
4.07
.007 .013 .013
3,350
11.69
8.11
.000 .003 .005
.20 −.12 .14 .14 −.16 −.16
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
79
Table 6 Results from multiple regression analyses (N = 351) on the influence of the clearness index (CI = 0.06–0.89), wind speed (w = 0.55–3.62 ms−1 ) and air temperature (Ta = −7.9 to 17.45 ◦ C) on the perceived “feelings right now” at the courtyard Dependent variable
R2
Glad–gloomy -w
.02
Calm–nervous -w
.02
Active–passive -CI -Ta
.08
Beta
df
MS
F
p
3,350
1.58
2.61
.051 .017
3,350
1.76
2.79
.040 .005
3,350
17.77
10.35
.000 .001 .001
−.13 −.16 .17 .19
Table 7 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the total attendance and perceived “current weather” in the park Dependent variable
R2
Total attendance -w -Ta
.36 .27
Calm–windy -w
.28
Beta
df
MS
F
p
3,56
517.21
9.71
.000 .020 .000
3,265
47.67
33.95
.000 .000
3,265
98.98
167.35
.000 .026 .000
3,265
10.66
12.57
.000 .005 .014 .020
.57 .53
Cold–warm -w -Ta
.66 .09 .82
Good–bad for outdoor activity -CI -w -Ta
.13 −.18 .16 −.14
The participants also assessed the current weather as being “good for outdoor activity” at times with clear skies, high air temperatures and low wind speeds.
the park as a “calm” and “warm” place. The results also showed that the park was evaluated as a “warm” place at higher air temperatures.
3.4.2. Question 2: “What is your perception of the place at this moment?” The results showed that the participants perceived the park to be a more beautiful place on occasions with clear skies and low air temperatures (see Table 8). Low wind speeds and clear skies were shown to have a significant influence on the perceptions of
3.4.3. Question 3: “How do you feel in this place at this moment?” As shown in Table 9, the participants felt more bored and gloomy at times when the sky was cloudy in the park. In addition, the participants felt more passive on occasions with high air temperature.
Table 8 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the perceived “place right now” in the park Dependent variable
R2
Ugly–beautiful -CI -Ta
.05
Windy–calm -CI -w
.26
Cold–warm -CI -w -Ta
.24
Beta
df
MS
F
3,265
4.04
4.85
.003 .051 .021
3,263
42.69
31.10
.000 .049 .000
3,264
35.88
27.76
.000 .000 .000 .033
.13 −.16 .12 −.45 .26 −.27 .12
p
80
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Table 9 Results from multiple regression analyses (N = 266) on the influence of the clearness index (CI = 0.13–0.91), wind speed (w = 0.45–3.92 ms−1 ) and air temperature (Ta = −3.6 to 20.87 ◦ C) on the perceived “feelings right now” in the park Dependent variable
R2
Elated–bored -CI
.03
Glad–gloomy -w
.05
Active–passive -Ta
.19
Comfort scale -w
.05
Beta
df
MS
F
p
3,263
2.51
2.78
.041 .007
3,265
2.48
4.32
.005 .008
3,264
33.41
19.88
.000 .000
3,159
2.98
−.18 −.18 .43 .17
2.68 .042
.049
Table 10 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the total attendance and perceived “current weather” at the waterfront plaza Dependent variable
R2
Total attendance -w -Ta
.22
Beta
df
MS
F
3,45
75.67
3.9
.015 .028 .004
3,201
45.51
36.73
.000 .000 .000
3,201
16.71
19.80
.000 .002 .000
3,201
14.80
15.89
.000 .000 .000
−.43 .56
Calm–windy -CI -w
.36
Cold–warm -w -Ta
.23
Good–bad for outdoor activity -CI -w
.19
−.23 .58 −.26 .60 −.41 .30
3.5. Results from study area 4: the waterfront plaza 3.5.1. Question 1: “What is your perception of the weather today?” The results showed that the number of participants at the waterfront plaza increased at times with lower wind speeds and higher temperatures (see Table 10). At this place, the participants perceived the current weather to be more windy, colder and worse for outdoor activity when the wind speed increased. Clearer skies, however, led the participants to assess the current weather as being calmer and better for outdoor activity. The weather was perceived to be warmer when the air temperature rose.
p
3.5.2. Question 2: “What is your perception of the place at this moment?” As shown in Table 11, the participants assessed the waterfront plaza as being more beautiful at higher wind speeds and lower air temperature. The waterfront plaza was also assessed as a “warmer” and “calmer” place on occasions with clear skies and when the wind speed increased. 3.5.3. Question 3: “How do you feel in this place at this moment?” The results showed that only the active–passive behavioural dimension was significantly influenced by the three weather variables (see Table 12). More precisely, the participants at the
Table 11 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the perceived “place right now” at the waterfront plaza Dependent variable
R2
Ugly–beautiful -w -Ta
.04
Windy–calm -CI -w
.21
Cold–warm -CI
.05
Beta
df
MS
F
3,201
2.82
2.98
.033 .004 .022
3,201
23.06
17.04
.000 .005 .000
3,201
3.81
3.23
.024 .008
.27 −.21 .19 −.43 .19
p
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
81
Table 12 Results from multiple regression analyses (N = 202) on the influence of the clearness index (CI = 0.09–0.84), wind speed (w = 0.42–9.47 ms−1 ) and air temperature (Ta = −2.89 to 16.91 ◦ C) on the perceived “feelings right now” at the waterfront plaza Dependent variable
R2
Active–passive -CI -w -Ta
.10
Beta
df
MS
F
p
3, 201
13.49
7.64
.000 .014 .030 .000
.18 −.20 .38
4.1. Weather parameters and the functional component
Fig. 4. The complex interrelation between weather and microclimate and the three components of place, adapted from Canter (1977).
waterfront plaza felt more active at high wind speeds, lower air temperatures and cloudy skies. 4. Discussion In general, the results of this study confirm the hypothesis that the air temperature, wind speed and clearness index (i.e., cloud cover) have a significant influence on the participants’ weather assessments and place-related perceptions, emotions and attendance. The results clearly show that weather and microclimate have a significant influence on two (functional and psychological) of three components constituting a place (Fig. 4).
Regarding the functional component of a place, measured as the participants’ attendance, the air temperature was shown to have a similarly significant influence in all places. Fig. 5 gives a schematic illustration of this impact (reported in Tables 2, 5, 8 and 11). It is clear from the diagrams in Fig. 5 that when the air temperature rose, the number of visitors increased in all four places. The statistical analysis not only showed that these results were significant, it also showed that 50% of the variance in attendance could be explained by weather parameters (CI, Ta , w), where Ta explains up to 63% of the variance (Table 4). Another general result for all four places was that the participants classified clear sky weather and weather with weak winds as being better for outdoor activity. A schematic illustration of this effect, based on results presented in Tables 2, 5, 8 and 11, is presented in Fig. 6. These results are in line with previous studies performed in Scandinavia (Carlestam, 1968; Gehl and Yencken, 1996; Thorsson, 2003; Thorsson et al., 2004) and other parts of Europe (Nikolopoulou and Steemers, 2003; Nikolopoulou and Lykoudis, 2006), which demonstrated a strong influence of weather on the use of outdoor urban spaces. The weather and climate in Scandinavia are characterised as both variable and changeable. It is also well-known that in these countries, the dark winter and the following short summer have fostered a special relationship between the inhabitants and the sun (Gehl and Yencken, 1996). If it is possible to enjoy the sun for only a short period, the urge to do so is tremendous. Results from a study that was conducted by the authors of this article in Matsudo, a satellite city of Tokyo, Japan, using the same methods as those presented above, showed a comparably low influence of weather on the use of different outdoor spaces (Thorsson et al., 2006). This difference between Sweden and Japan is probably due to both cultural and climatological differences (Knez
Fig. 5. Total attendance as a function of Ta , CI and w (square, courtyard, park and waterfront plaza). The diagrams are a schematic illustration of the statistics reported in Tables 2, 5, 8 and 11.
82
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Fig. 6. Plots of good–bad settings for outdoor activity as a function of Ta , CI and w (square, courtyard, park and waterfront plaza). The diagrams are a schematic illustration of the statistics reported in Tables 2, 5, 8 and 11.
and Thorsson, 2007). Large differences in comfort conditions between different cities in Europe were also reported from the RUROS project (Nikolopoulou and Lykoudis, 2006). 4.2. Weather parameters and the psychological component It is known that architects, urban planners and climatologists have long argued that climate-related issues are important for city life. But how do the people living in the city, i.e., “the public”, relate to climatic elements, and are weather and climate of any importance to an individual (Eliasson, 2000)? Based on the results presented in this study, this question must be answered in the affirmative. It is very clear from the study that the people visiting the four places do care about, and are influenced by, the weather and climate. Thus, weather parameters are important—not only did they influence the number of people who visited the spaces and whether they assessed the weather as being good or bad for outdoor activity, but they also influenced how people assessed the places and how being in these places made them feel, i.e., the psychological component of a place. Concerning the psychologically-related results, and as pointed out in Section 3, it must be noted that only tentative links between participants’ behaviour, weather and place were indicated, due to the low R2 values. This type of result is related to measurement errors in subjective assessment and to the large body of uncontrolled variables that may also have had an impact on participants’ psychological responses. An interesting result is that the open, exposed waterfront plaza was estimated to be more beautiful as a result of high wind speed and low air temperatures, as shown in Fig. 7. Participants also felt more active when there was higher wind speed in the waterfront plaza (Table 2). In contrast, participants visiting the open square estimated it to be more beautiful as a result of lower wind speeds and
Fig. 7. Ugly–beautiful as a function of w at the square (left diagram) and waterfront plaza (right diagram). The diagrams are a schematic illustration of the statistics reported in Tables 3, 6, 9 and 12.
higher air temperatures. A tentative explanation is that wind at the waterfront renders a positive aesthetic and symbolic value. This place gains life through the natural elements reflected in the water and the activities at the adjacent marina become more interesting to look at on a windy day. The attraction of the square, and to some extent even the park, are other people. The typical park or square activities are merely reduced by the wind. This corresponds with the theory put forward by Westerberg (1994), who illustrated that gustiness caused by unsuitable building geometry was disliked much more than the blustery winds on hilltops or natural maritime settings, where even “bad” weather would become an attraction. Nikolopoulou and Steemers (2003) arrive at the same conclusions: in urban spaces with high amounts of natural characteristics such as parks, the tolerance to widespread changes in the physical environment is high, provided they are produced naturally. Several other studies show that the experience of nature in the city is a source of positive feelings (e.g., Chiesura, 2004). The results also show that feelings of pleasantness increase with higher clearness index. Clear skies also increased the subjective weather-related assessments of warmth and calm in three of the four urban spaces (square, park and waterfront plaza, Table 1). 4.3. Weather parameters and the physical component Weather and climate influence the physical component of a place. However, as illustrated in Fig. 4, the relationships also work in the opposite direction. A building’s shape, orientation, material, colour, etc., influence radiation, temperature, wind and other parameters, to produce a site-specific microclimate. Urban designers/planners are thus able to create urban environments that take advantage of the positive effects of the existing weather and climate. Climate design has always been a natural part of local building tradition. However, when designing the modern city, the urban designer/planner has to consider many different conflicting aspects. Thus climate-related issues have a relatively small impact on the planning process (Eliasson, 2000). We hope, however, that studies such as this one will serve to bolster the arguments in favour of using climate-sensitive planning in future designs of cities. What are these arguments? Usage and activity are often used as measures of successful urban spaces. For example, Carmona et al. (2003, p. 99) states that “successful public places are characterised by the presence of people”. The present study shows that in order to create successful public places, weather parameters cannot
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
be neglected, since they account for over 50% of the variance in place-related attendance. Viewed together with results illustrating the influence of weather parameters on perceptions and emotions, it is clear that climate-sensitive planning can be an important tool in the drive to increase sustainability in the urban environment. Attractive urban spaces influence the social life in the city and will also indirectly influence the local economy and transportation infrastructure. In this perspective, climatesensitive planning may influence all of the three dimensions of sustainability, i.e., environmental, social and economic qualities. More research on the human response to urban climate is however needed before this theory can be proved. The compact city versus an urban sprawl is one of the key issues facing future urban sustainable development, both in Sweden and abroad. It is important to remember that climate-sensitive planning of urban public places is important, regardless of which future direction is chosen. In practice, climate-sensitive planning is a matter of choosing between alternatives. It is difficult to apply one general rule to all types of urban spaces. Regardless of whether it is a new construction or a re-design, a starting point for climate-sensitive planning is to take advantage of the seasonal variations of the microclimate. Urban spaces can be “climatically” attractive all year round. One way is to design a differentiated microclimate within the place. As each urban space is unique, it is important to first identify environmental characteristics such as openness, shadow pattern and wind conditions within the place and then simulate the proposed changes. Several interesting projects have recently developed GIS-based models as tools for analysing environmental diversity (e.g. Ratti and Richens, 2004; Steemers, 2006). 5. Conclusion Weather parameters (clearness index, air temperature and wind) have a significant influence on participants’ weather assessments and place-related perceptions, emotions, attendance. It is thus clear that solar radiation, air temperature and wind are vital aspects of the functional and psychological components of a place. The results support the arguments in favour of employing climate-sensitive planning in future urban design and planning projects, as the physical component of a place can be designed to influence the site-specific microclimate and consequently people’s place-related attendance, perceptions and emotions. Acknowledgements The project is financially supported by Formas, the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning in their key action area Urban Public Spaces. References Blaney, P.H., 1986. Affect and memory: a review. Psychol. Bull. 99, 229–246. Canter, D., 1977. The Psychology of Place. The Architectural Press Ltd., London.
83
Canter, D., 1997. The facets of place. In: Moore, G.T., Marams, R.W. (Eds.), Advances in Environment, Behavior, and Design. Plenum press, New York, pp. 109–147. Carlestam, G., 1968. Studier av utomhusaktiviteter med automatisk kamera. Rapport R16:68, Byggforskningsr˚adet. Carmona, M., Heat, T., Oc, T., Tiedsell, S., 2003. Public Places—Urban Spaces: The Dimensions of Urban Design. Archtectural Press, Elsevier. Chiesura, A., 2004. The role of urban parks for the sustainable city. Landsc. Urban Planning 68, 129–138. Cohn, E.G., 1993. The prediction of police calls for service: the influence of weather and temporal variables on rape and domestic violence. J. Environ. Psychol. 13, 71–83. Eliasson, I., 2000. The use of climate knowledge in urban planning. Landsc. Urban Planning 48, 31–44. Gehl, J., 1971. Livet mellem husene. K¨openhamn. Gehl, J., Yencken, D., 1996. Byens rum—byens liv. Arkitektens forlag. Kunstakademiets forlag. Gifford, R., 1980. Environmental dispositions and the evaluation of architectural interiors. J. Res. Pers. 14, 386–399. Graumann, C.F., 2002. The phenomenological approach to people-environment studies. In: Bechtel, R.B., Churchman, A. (Eds.), Handbook of Environmental Psychology. John Wiley & Sons, Inc., New York, pp. 95–113. H¨oppe, P., Seidl, H.A.J., 1991. Problems in assessment of the bioclimate for vacationists at the seaside. Int. J. Biometeorol 35, 107–110. Knez, I., 2005. Attachment and identity as related to a place and its perceived climate. J. Environ. Psychol. 25, 207–218. Knez, I., 2006. Autobiographical memories for places. Memory 14, 359–377. Knez, I., Hygge, S., 2001. The circumplex structure of affect: a Swedish version. Scand. J. Psychol. 42, 389–398. Knez, I., Thorsson, S., 2006. Influence of culture and environmental attitude on thermal, emotional and perceptual evaluations of a square. Int. J. Biometeorol 50 (5), 258–268. Knez, I., Thorsson, S., 2007. Thermal, emotional and perceptual evaluations of a park: cross-cultural and environmental attitude comparisons. Environ. Behav., submitted for publication. Kuiken, D., 1991. Mood and Memory: Theory, Research and Applications. Sage, London. Lindberg, F., 2005. Towards the use of local governmental 3-D data within urban climatology studies. Mapping Image Sci. 2005 (2), 4–9. Matzarakis, A., Mayer, H., 1996. Another kind of environmental stress. WHO News 18, 7–10. Mayer, H., H¨oppe, P., 1987. Thermal comfort of man in different urban environments. Theor. Appl. Clim. 38, 43–49. Mills, G., 1999. Urban Climatology and Urban Design. ICUC, International Conference on Urban Climatology in Sydney, Australia, 8–12 November. Extended abstracts. Mills, G., 2006. Progress toward sustainable settlements: a role for urban climatology. Theor. Appl. Climatol. 84, 69–76. Nikolopoulou, M., Baker, N., Steemers, K., 2001. Thermal comfort in outdoor urban spaces: understanding the human parameter. Solar Energy 70 (3), 227–235. Nikolopoulou, M., Steemers, K., 2003. Thermal comfort and psychological adaptation as a guide for designing urban spaces. Energy Buildings 35, 95–101. Nikolopoulou, M., Lykoudis, S., 2006. Thermal comfort in outdoor urban spaces: analysis across different European countries. Building Environ. 41, 1455–1470. Parker, P.M., 1995. Climatic Effects on Individual, Social and Economic Behavior. Greenwood Press, Westport, CT. Pfaffenberger, R.C., Patterson, J.H., 1987. Statistical Methods. Irwin, Homewood, IL. Ratti, C., Richens, P., 2004. Raster analysis of urban form. Environ. Planning B Planning Des. 31 (2), 297–306. Rotton, J., Cohn, E., 2002. Climate, weather and crime. In: Bechtel, R.B., Churchman, A. (Eds.), Handbook of Environmental Psychology. John Wiley & Sons, Inc., New York, pp. 481–498. Simister, J., Cooper, C., 2005. Thermal stress in the U.S.A. effects on violence and on employee behaviour. Stress Health 21, 3–15.
84
I. Eliasson et al. / Landscape and Urban Planning 82 (2007) 72–84
Steemers, K., 2006. Human comfort in urban spaces. In: The 6th International Conference on Urban Climate (ICUC6). Thorsson, S., 2003. Climate, air-quality and thermal comfort in the urban environment. Doctoral Thesis A87. G¨oteborg University, Sweden. Thorsson, S., Lindqvist, M., Lindqvist, S., 2004. Thermal bioclimatic conditions and patterns of behaviour in an urban park in Sweden. Int. J. Biometeorol. 48, 149–156. Thorsson, S., Honjo, T., Lindberg, F., Eliasson, I., Eun-Mi, L., 2006. Thermal comfort and outdoor activity in Japanese urban public spaces. Environ. Behav., in press. Tress, G., Tress, B., Fry, G., 2004. Clarifying integrative research concepts in landscape ecology. Landsc. Ecol. 20, 479–493. Westerberg, U., 1994. Climatic planning—physics or symbolism. Architecture Behav. 19, 49–72. Zacharias, J., Stathopoulos, T., Wu, H., 2001. Microclimate and downtown open space activity. Environ. Behav. 33, 296–315. Ingeg¨ard Eliasson Born on 1961, presented her PhD thesis in 1993 and is since 2000 associate professor in physical geography at G¨oteborg University. Her current research focus on studies of wind and turbulence field and energy balance in urban street canyons, human biometeorology and thermal comfort, studies of urban climate and air pollution in African cities and integration of climate knowledge in urban planning.
Igor Knez Born on 1959, has, since 1992, a PhD in psychology from the University of Uppsala, Sweden. Since 1998, he is associate professor in psychology at Centre for Built Environment, University of G¨avle, Sweden. His current environmental psychology research focuses on place-related identity and memory and place-related psychological assessment. Ulla Westerberg Born on 1946, received her PhD in 1993 on climatic design in architecture. She has been the project leader of urban climate and housing surveys and has cooperated in projects on energy efficiency and environmental assessment of buildings at Centre for Built Environment, G¨avle University. Sofia Thorsson Born on 1972, received her PhD in 2003 in physical geography. She went on post-doc at Chiba University, Japan and University of G¨avle, in 2004 and 2005. Since 2005 she is post-doctoral Fellow at G¨oteborg University. Research focus is on urban climate, human biometeorology, thermal comfort and air pollution in urban areas. Fredrik Lindberg Born on 1974, received his MA at G¨oteborg University in 2002 and is currently a PhD-candidate in physical geography at G¨oteborg University, Sweden. His main expertises are urban climatology, human biometeorology and geographical information science.