Human exposure to environmental health concern by types of urban environment: The case of Tel Aviv

Environmental Pollution 208 (2016) 58e65

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Human exposure to environmental health concern by types of urban environment: The case of Tel Aviv Izhak Schnell a, Oded Potchter a, b, *, Yaron Yaakov a, Yoram Epstein c a

Department of Geography and Human Environment, Tel Aviv University, Tel Aviv, Israel Department of Geography, Beit Berl Academic College, Beit Berl, Israel c Heller Institute of Medical Research, Sheba Medical Center, Tel Hashomer and The Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 May 2015 Received in revised form 14 August 2015 Accepted 21 August 2015 Available online 4 September 2015

This study classifies urban environments into types characterized by different exposure to environmental risk factors measured by general sense of discomfort and Heart Rate Variability (HRV). We hypothesize that a set of environmental factors (micro-climatic, CO, noise and individual heart rate) that were measured simultaneously in random locations can provide a better understanding of the distribution of human exposure to environmental loads throughout the urban space than results calculated based on measurements from close fixed stations. We measured micro-climatic and thermal load, CO and noise, individual Heart Rate, Subjective Social Load and Sense of Discomfort (SD) were tested by questionnaire survey. The results demonstrate significant differences in exposure to environmental factors among 8 types of urban environments. It appears that noise and social load are the more significant environmental factors to enhance health risks and general sense of discomfort. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Measurements techniques Urban micro-environments Personal exposure Thermal comfort Air pollution Noise Social stress

1. Introduction Over the last decade the combined effects of a set of environmental factors on subjective sense of discomfort and health concern have received growing research attention, given mounting awareness of the risks posed by urban heat islands, concentrations of air pollution, noise, visual and social loads, and similar phenomena (Ulrich et al., 1991b; Peschardt and Stigsdotter, 2013). Studies in this field tend to define as dependent variables either their subjects' accounts of subjective sense of discomfort (SD) (Toftum, 2002; Fang et al., 2004; Schnell et al., 2012) or physiological indices such as Heart Rate Variability (HRV) (Schnell et al., 2013), Salivary Cortisol (Van den Berg and Custers, 2011) and cognitive tests (Kaplan, 1995). Most studies have focused on the effects of one or two environmental factors on health. Several studies focused on the effects of thermal loads and noise on subjective sense of discomfort (Epstein et al., 2000; Toftum, 2002; Pellerin and Candas, 2004;

* Corresponding author. Department of Geography and Human Environment, Tel Aviv University, Tel Aviv, Israel. E-mail address: [email protected] (O. Potchter). http://dx.doi.org/10.1016/j.envpol.2015.08.040 0269-7491/© 2015 Elsevier Ltd. All rights reserved.

Candas and Dufour, 2005). Other studies have focused on the effects of thermal load and air pollution on subjective sense of discomfort (Poupkou et al., 2011). Recently, more comprehensive studies are considering four factors (CO, thermal load, social load and noise) (Schnell et al., 2012, 2013). While most studies focused on the effects of environmental factors on subjective sense of discomfort, a growing number of studies measure the effects of environmental factors on HRV (Sawasaki et al., 2001; Kurosawa et al., 2007; Bjor et al., 2007; Liu et al., 2008; Rashid and Zimring, 2008; Schnell et al., 2013). All of the studies noted above found significant effects of all these environmental factors on risk for health and subjective sense of discomfort. In studying variations in concentrations of environmental risk factors within indoor and outdoor urban types of environments, most studies use mathematical models based on measurements extracted from a small number of fixed climatic monitoring stations distributed around the city and extrapolations of these results based on few factors like distance from pollution sources, volume of risk at the source and intervening variables that affect patterns of distribution like wind direction and velocity, urban morphology, etc (Gu et al., 2011; Sun et al., 2012). Such models are vulnerable to several critiques. First, numerous studies show that monitoring stations located in cities tend to underestimate concentrations of

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

pollutants to which residents are exposed in practice (Duci et al., 2003; Gullver and Briggs, 2003; Kaur et al., 2005; Potchter et al., 2014). Second, studies show that the pollutant distribution patterns in urban spaces are highly complex, to an extent that makes it difficult to accurately model them (Zwack et al., 2011). Third, using data from fixed outdoor monitoring stations does not enable comparison between indoor and outdoor environments. Previous studies have focused on the differences in subjective sense of discomfort and health risks that different types of environments pose. Such studies aim to expose the environmental risk factors in inner city environments while highlighting the restorative power of green areas (Ulrich et al., 1991a; Kaplan, 1995; Kaplan and Kaplan, 1989; Hartig et al., 2001; Parsons et al., 1998; Korpela et al., 2008; Staats et al., 2008). However, they suffer from three major shortages. First, they focus almost exclusively on the contrast between the busiest central urban environments and the green environments at cities' outskirts, ignoring the ‘mid-urban’ environments in which most people reside and act. We study diverse human exposures to environmetal discomforts and health risks in different types of urban environments. Second, since other studies measure exposure to environmental discomforts and health risks without controlling the concentrations of environmental risk factors in the measured sites, we added simultaneous measurements of environmental risk factors (thermal load, air pollution, noise and social load) that might affect the subjective sense of discomfort and health risk. Third, no agreed upon measurements or associations exist between subjective sense of discomfort and HRV; former studies tend to apply the more frequently used measures, with the exception of Park et al. (2010) and Schnell et al. (2013), which show discrepancies among the results for subjective psychological and physiological measurements. We suggest comparing the results for both subjective sense of discomfort and HRV and exposing the causes for differences between them. Schnell et al. (2012; 2013) offered a new methodology to monitor the effects of a set of environmental factors on subjective sense of discomfort and health, by HRV. In these studies, individuals' exposure to environmental variables such as noise, CO and micro climate as well as individuals' heart pulse rates were monitored by mobile micro-sensors carried on the subjects' bodies while they were performing their daily life in the city. This quantitative assessment was followed by questionnaires assessing the subjects' social loads and subjective senses of discomfort. This study aims to classify urban environments in which measurements were taken into types characterized by different combinations of exposure to environmental risk factors and to measure average levels of subjective sence of discomfort and HRV produced in each of these types of environments. We hypothesize that a set of measurements taken by the same individuals in different types of urban environments in a random way can provide a better understanding of the distribution of human exposure to environmental risk factors throughout the urban space than results calculated based on measurements from close fixed stations. 2. Research methods 2.1. Methodological approach This study adopts an urban ecological approach based on three domains: 1. Spaces investigated: urban environments that are frequented by those ascribing to the young 'Urbanite' socio-spatial lifestyle. 2. Independent variables: exposure to noise, CO, thermal load and social stress as a set of environmental factors. 3. Dependent variables: HRV and subjective sense of discomfort (SD) as the results of this set of environmental factors. These independent variables were chosen since the human body responds

59

immediately to these factors and this response can be seen in the heart rate variability (Schnell et al., 2013). 2.2. Study area The study was conducted in the city of Tel Aviv, Israel, located at 32
060 N 34
470 E, situated along the eastern coast of the Mediterranean Sea. The climate of Tel Aviv is defined as Subtropical € ppen classification Mediterranean, Csa according to the Ko (Potchter and Saaroni, 1998), characterized by a mild, wet winter and a hot, humid, summer (Bitan and Rubin, 1994). Tel Aviv, the core of the largest metropolis of Israel, with population of 3.46 million, has a population of 414,600 (Statistical Abstract, 2012).Tel Aviv is a modern city that passes an accelerating growth over the last 100 years, demonstrates the activity phenomenon that characterizes any modern city. 2.3. Research subjects and experimental course Thirty-six subjects participated in the study (22 males and 14 females). The subjects were healthy people between the ages of 23 and 40, who avoided smoking throughout the course of the experiment and 72 h prior to the experiment. None of them were regular users of drugs and alcohol or any medication. The subjects followed a pre-determined daily route starting from the university in northern Tel Aviv to Jaffa at the southern end of the city and back. Each experiment lasted for two sequential days, including six subjects following the same route. Two such experiments took place during the winter, two during the summer, one during the autumn and one during the spring. All 36 subjects followed the same route. Along the route, the subjects stopped at approximately 10 sites every day. The investigated sites included indoor sites such as shopping malls, student dormitories or apartments and pubs; as well as outdoor areas, including open markets, main streets, side streets and parks. Between the sites, participants used busses for longer distances and walking for shorter ones. The subjects spent about 45 min in each site; they spent at least 15 min at each site, adjusting, before measurements were taken (Fig. 1). 2.4. Measurements techniques of personal exposure to independent set of environmental factors In order to evaluate the personal exposure to an independent set of environmental factors in the micro urban environment, four methods were used: (1) measurements of micro-climatic and micro-environmental variables, (2) calculations of thermal sensation, (3) calculation of Heart Rate Variability (HRV) and (4) a questionnaire survey. Three fix environmental monitoring stations of the Israeli ministry of environmental protection are situated along the predetermined daily route and used as a background information data. The subjects wore a suit with portable measuring equipment that measured in-situ climatic and environmental variables, using mobile micro-sensors that continuously measured noise and CO levels, as well as climatic variables. CO Levels were measured and recorded using the portable Drager Pac III logger with CO sensor once every 60 s throughout the experiment. The instruments were calibrated both with 25 ppm CO calibration gas and fresh air before each experiment, and compared to CO data collected by one of the monitoring stations of the Israeli Ministry for Protection of the Environment. The correlation coefficient (R) was determined to be 0.79, with the Pac III more sensitive to immediate changes than the Ministry's recordings. Noise measurements were recorded with a Quest pro DL dosimeter ranging from 40 to 110 dB, with resolution of 0.1 dB. The noise sensors were

60

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

Fig. 1. Study area and the subjects' route of movement during the measurements.

calibrated before each experiment using a QC-10 calibrator (114 dB, 1000 Hz). We measured the average noise level per minute over the course of the experiment. A Fourier Microlog with a resolution of 0.5
C and accuracy of ±0.6
C measured climatic variables of temperature and relative humidity. Relative humidity was measured in each measurement site with a resolution of 0.5% and an accuracy of ±3%. Radiant temperature and cloud cover were taken from the Israeli Meteorological Service and calculated according to subjects' sun exposure. Wind velocity was using the Kestrel 3000 environmental monitoring hand held system. Thermal load was calculated using the Physiological Equivalent €ppe, 1999) which is the most Temperature (PET) index (Ho

commonly applied index for evaluation of human thermal perception and has been tested by questionnaire surveys in many s et al., 2006; Johansson and Emmanuel, 2006; field studies (Gulya Matzarakis et al., 2007; Lin, 2009; Cohen et al., 2013, 2014). PET was calculated using RayMan modeling program (Matzarakis et al., 2007) developed according to Guideline 3787 of the German Engineering Society (VDI, 1998), For a man with body surface area of 1.9 m2, a height of 1.75 m and a body weight of 75 kg, metabolic rate of 80 W/m2 for a standing person, insulation factor of clothing 0.9 € ppe, 2002). (Ho A Polar 810i monitor recorded heart beat intervals (ReR) for each subject continually over the entire course of the experiment. The ReR signal was converted to heart rate, linearly interpolated

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

and graphed in regular intervals of 0.5 s (2 Hz). Segments of seven to ten minutes, at a steady state condition, were obtained and analyzed for each subject at each site along the way. We applied the frequency domain index of HRV. In addition to total power, two different frequencies were analyzed: low-frequency power (LF) (0.04e0.15 Hz) and high-frequency power (HF) (0.15e0.4 Hz). According to the North American Society of Pacing and Electrophysiology and the Task Force of the European Society of Cardiology, the HF domain dimly represents mediated respiratory variations, whereas the LF domain represents predominantly sympathetic activity (Hainsworth, 1995). In order to highlight the controlled and balanced behavior of the two branches of the autonomic nervous system, LF and HF were calculated in normalized units and the LFeHF ratio was calculated as a measure of autonomic balance LF/ HF (Wang et al., 2005). Simultaneously with the empirical data collection, subjects filled out a questionnaire survey after they stayed in a stable state for 10 min in the middle of the 30e45 min time interval allocated for their visit to each of the stations, as well as in the middle of the route between sequential places, thereby producing about 20 survey responses per day per participant. These survey responses provided data pertaining to the environmental and climatological situation and personal HRV. In addition, questionnaire responses helped evaluate the Subjective Social Load and Sense of Discomfort. The answer to the question, “to what extent are you stressed by the presence of other people in your immediate surroundings? And “to what extent do you feel comfort or discomfort in this environment?” (on a grade from one to ten) were applied towards assessing social load and subjective sense of discomfort respectively.

3. Analysis The analysis was performed in three stages. The first stage involved aggregating the 698 measurements into types of environments and analyzing their internal distributions. This process led to the calculation of mean levels of HRV and SD for each type of environment. Next, we have calculated the significance of differences in levels of risk for health among the types of places base on two tailed repeated measure ANOVA. Third, we calculated mixed models accounting for repeated measures and interactions among the variables in order to measure association between a set of environmental risk factors and types of places on the one hand and risk for psychological and physiological measurements of risk for health. We categorized six types of environments, with over 50 measurements of all environmental, climatological, personal HRV and questionnaire responses applicable towards each type. This characterization enables comparison between indoor and outdoor environments, more or less crowded ones, and environments characterized by different environmental settings (Table 1).

61

4. Results In terms of SD, the six types of environments can be aggregated into three groups according to their means and standard deviations values (Table 2). The most risky environments are the students' dormitories. In this type of environment, subjects suffered from both noise and high exposure to other people. On the other end of the spectrum, the more relaxing environments were parks. This finding is not surprising: parks are perceived as the main outdoor relaxation sites in cities (Ulrich et al., 1991a; Korpela et al., 2008; Tsunetsugu et al., 2013). Next in terms of level of subjective comfort experienced are the main streets e Tel Aviv's Rothschild and Jerusalem Boulevards, characterized by a large volume of greenery. The third group of environments includes side streets, shopping malls and open markets, which subjects experienced as relatively uncomfortable. In terms of LF/HF, the six types of environments can be divided into two groups (Table 2). The more populated places of shopping malls, open markets and main streets were experienced as more risky, while the other environments were experienced as least risky. Surprisingly enough, dormitories that were evaluated as a risky environment in terms of SD, were characterized by low levels of LF/HF. As in the case of SD parks appeared to be the more relaxing outdoor environment in terms of LF/HF. Notably, SD and LF/HF measure different aspects of human response to environmental risk factors. Pearson correlation coefficient between them shows that variability in SD predicts only eight percent of the variability in LF/HF, although the correlation is highly significant (R ¼ 0.28; T ¼ 7.1; sig. 0.0001). Some significant differences in ranking types of environments in terms of SD and LF/HF also emerged. Especially deep gaps, however, were measured in dormitories and side streets, in which SD levels were much higher than would have been expected according to their respective LF/HF values and exceptionally low discrepancies between these different indicators were measured in parks, which were experienced as relaxing according to both indices (Table 2). 4.1. Differences between types of environments Our second stage of analysis investigated the significance of the differences in exposure to subjective discomfort and health risks between the different types of environments (In this analysis we include also pubs and we distinguish between students dormitories and students who sleep at home with their parents). Tables 3 and 4 present the results for two tailed repeated measure ANOVA F-Test calculations for subjective sense of discomfort and LF/HF. The analyses for subjective discomfort show that 19 out of the 29 possible relations are significant at a level of <0.05, and eight of them are significant at levels of <0.001 (Table 3). While differences among main streets, side streets, open markets and shopping malls remained insignificant, most of the differences among the other types of environments proved to be significant. Pubs appear to

Table 2 Mean levels of SD and LF/HF by type of place.

Table 1 Number of measurements by type of place. Type of place

Number of measurements

Main st. Side st. Open markets Shopping malls Urban parks Dormitories Apartments Total

102 50 52 187 76 54 177 698

Type of place

SD Mean

s.d.

Mean

s.d.

Main st. Side st. Open market Shopping mall Park Dormitories

28.6 33.7 30.2 30.3 26.1 40.7

19.3 20.1 19.8 16.1 16.2 27.8

12.4 11.1 14.5 13.5 10.3 8.1

11.2 10.1 10.9 11.6 9.1 5.7

Notes: 1. 2. F ¼ 8.3; Sig. ¼ 0.0001; df ¼ 11.

LF/HF

62

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

Table 3 Two tailed repeated measure ANOVA between types of places and sense of discomfort.

Table 4 Two tailed repeated measure ANOVA between types of environments and LF/HF.

present the most extreme exposure to subjective sense of discomfort e far beyond any other type of environment except for the students' dormitories, which seem to provide similar levels of environmetal discomfort to pubs. Apartments differ from all other environments due to the exceptionally low levels of environmetal discomfort they produce. In between are parks and main tree-lined streets, which exposed the subjects to significantly lower levels of environmetal discomfort than did side streets, open markets, shopping malls, dormitories and pubs, but to significantly higher levels of environmetal discomfort relative to apartments. The analyses for LF/HF show that only 15 of the 29 possible relations are significant at levels of <0.05 and only two of these at levels of <0.001 (Table 4). In terms of HRV, the dormitories pose the lowest risk for the subjects' health, mainly in comparison to main streets, open markets and shopping malls. Parks offer significantly healthier environments than side streets, shopping malls, open markets and pubs, in which the levels of health risks are similar, but significantly less healthy environments than do dormitories. 4.2. The effects of environmental factors on subjective sense of discomfort and HRV In order to analyze the effects of the studied environmental factors on SD and LF/HF, we calculated two analyses of mixed

models: one for each dependent variable (Table 5). In the analyses we calculate the effects of the environmental risk factors on LF/HF and subjective sense of discomfort. We add to it the factor of types of place in order to investigate whether types of places pose any effect on risk for health independently of the other environmental risk factors. The first conclusion that appears from Table 5 is that noise and social load are the more dominant environmental risk factors in the case of Tel Aviv. Their dominance is somewhat more prominent in terms of their effects on LF/HF

Table 5 Mixed models for associations between a set of environmental risk factors and types of places on the one hand and risk for health on the other hand. Environmental factor

Sense of comfort (SD) F

Place type 3.9 Thermal load (PET2ln) 9.2 Social load (ln) 5.6 Noise (ln) 4.5 CO (ln) 0.12 Thermal e load*social load Thermal load*noise e Social load*noise e

T

Sig

1.5 3.7 2.4 2.4 0.3

0.001 0.003 0.019 0.036 0.73 e e e

b

HRV (LF/HF) F

T

0.30 2.6 0.8 0.43 1.8 1.3 0.30 9.4 2.9 0.30 25.4 5.4 0.04 0.1 0.2 14.5 3.8

b

Sig 0.67 0.097 0.004 0.0001 0.986 0.0001

12.3 3.5 0.001 9.9 3.2 0.002

0.21 0.07 0.31 0.38 0.01

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

(b ¼ 0.38 and 0.36 respectively) than on subjective sense of discomfort (b ¼ 3.0) and 3.0 respectively in which thermal load is the more dominant environmental risk factor (b ¼ 0.43). Furthermore, social load and noise interact with each other in effecting LF/ HF. This result hints at the fact that significant part of the noises in Tel Aviv are caused by human beings in public spaces. Thermal load is the most prominent environmental risk factor in effecting subjective sense of discomfort both in extreme hot and cold weathers. The effects of thermal load influence mainly subjective sense of discomfort but also indirectly LF/HF (Table 5). Thermal load interacts with noise and social load in order to affect LF/HF. It seems that in extreme weather conditions, when thermal load is associated with noise and social load, human beings lose coping capabilities in a way that increase their risk for health as it is measured by LF/HF. However, some variations exist in the contribution of each environmental factor to the explanation of variations in SD and LF/ HF. Three environmental risk factors affect SD primarily: thermal load; noise; and social load and type of place levels affect it secondarily (Table 5). Higher concentrations of the three primary factors contribute to increase in subjective sense of discomfort and mainly parks contribute to reduced levels of SD independently of the effect of the environmental risk factors. The negligible levels of CO in all environments left CO with no significant effects on SD (Table 5). In addition, no interactions among the independent variables of the environmental risk factors were recorded. Noise, social load and thermal load also significantly affect LF/HF including the interactions among them (Table 5). Like in the case of SD the negligible levels of Co do not have any significant effect on LF/HF. Most interesting, despite the fact that half of the differences among levels of LF/HF among the types of places are insignificant (Table 4) the influence of type of place remain in the mixed model significant though with lower influence than on subjective sense of discomfort (b ¼ 0.21). This means that types of places influence LF/ HF independently of the aforementioned environmental factors: thermal load, Social load and noise. The last analysis focuses on the power of the mixed models to explain the health risks and stress posed by the environmental factors which were studied and the types of places. It turns out that the four environmental factors explain 57 percent of the variability of DS and 75 percent of the aforementioned environmental factors and the types of places. This means that other factors associated with types of places explain 18 percent of the variability in subjective sense of discomfort. The same four environmental factors explain 35 percent of the variability of LF/HF and 41 percent of the combined influence of the four aforementioned environmental factors and types of places. This means that in terms of LF/HF other characteristics of types of places explain only 6 percent of the variability in LF/HF. 5. Discussion and conclusions This article suggests a new methodology for the analysis of human exposure to risks for health in urban areas. Current studies are based on measurements in few fixed stations and models that evaluate distributions of risk effects in urban spaces. This study is based on empirical data extracted from peoples' bodies while and where they perform their everyday life in the city. Random samples of such measurements in each type of environment have been taken supplying the study with empirical real data that replace the simulated ones from fixed stations. Inaccuracies may stems from the tendency of fixed stations to undermine the real levels of human exposure to environmental risk factors, the oversimplification of the simulation models and the lack of sensitivity to spatial variations in human exposure (Schnell et al., 2013). We argue that the

63

empirical methodology suggested in this study for assessing the impact of environmental risk factors on people's sense of discomfort and risk for health by using micro-sensor has advantage over the present classic monitoring approach used by the municipally and governmental ministries. However, this integrated methodology needs to be expanded on larger number of subjects, in diverse urban environments and broaden the investigated population to different age groups (infants, children's and elderly). From a technical point of view, it seems that there is a need to produce portable micro devices in order to increase the span of air pollutants accurate measurements. In interpreting the results of the analysis of a study performed in Tel Aviv we argue that it is widely accepted that urban environments are highly risky in terms of both subjective sense of discomfort and risk for health, as measured by HRV (Schnell et al., 2012, 2013). However, the relations between these factors have not been investigated. This study reveals significant and positive correlations between SD and LF/HF, although these correlations are relatively low. These differences appear prominent in comparing levels of risk experienced in the different types of environments. Meaningful discrepancies (above two environments in the ranking of the two scales) emerged in four of the eight types of environments or in three of the six environments characterized here. The deepest gap between levels of risk in types of environments was measured in dormitories, in which low levels of HRV followed a strong subjective sense of discomfort. A significant gap was recorded also in the case of pubs, in which the highest levels of subjective sense of discomfort were recorded with only moderate levels of HRV. On the other end of the spectrum, subjects experienced high levels of HRV and medium levels of subjective sense of discomfort in open markets and main streets. Tel Aviv residents perceive both of these environments to be attractive visitation spots due to the interest found in the market and the strip of trees along the main boulevards, despite the fact that both types of environments are highly noisy and crowded. In the other types of environments, similar levels of risk were experienced. Side streets and shopping malls provided moderate levels of risk, while the lowest levels of risks were experienced in parks. Parks seem to succeed in fulfilling their role of offering relaxation spaces that reduce health risks (in relation to other urban spaces). It seems that the cultural association of visits to parks and crowded markets with pleasure affects SD slightly more than HRV. The therapeutic power of green areas has been studied increasingly over the last decades, with accumulating evidence available for their restorative power (Kaplan, 1995; Ulrich et al., 1991a; Korpela et al., 2008). The present study aligns with this body of literature, but it adds more evidence of a more specific nature regarding the restorative power of parks. Most studies to date have compared human exposure to risks in the busiest downtown areas and in larger forests at the outskirts of the city (Peschardt and Stigsdotter, 2013; Brown et al., 2013; Tsunetsugu et al., 2013). Our study compares risks people encounter over the course of their daily lives in small urban parks in relation to a range of other typical urban environments. Even under these circumstances, parks supply more relaxing environments than any other outdoor environments or crowded indoor environments like shopping malls. Our findings suggest that a short rest in an inner city park during routine working days provides an effective means for relaxation and recovery. The study also shows that a tree-lined urban boulevard positively affects human sense of comfort. The study uncovers not only significant differences in exposure to health risks in parks as opposed to other inner city environments, but also differences in this domain among the other types of environments identified. In particular, the study highlights the risk

64

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65

effects posed by noisy and crowded environments of shopping malls and open markets. The crowded environments of shopping malls and open markets showed the highest levels of exposure to LF/HF, while visitors there noted relatively high levels of subjective sence of discomfort. Notably, while shopping malls' design generally aims to provide visitors with comfortable environments (Gilboa, 2006), it seems that the crowding and loud noises in malls render them almost as risky as open markets. Indeed, this phenomenon might be culture-specific: a study on shopping malls in Israel notes the noisy behavior of Israelis as a problem that designers are dealing with (Wang et al., 2005). This study confirms the negative effects of noise and crowding on visitors' sense of comfort and LF/HF. While visitors perceive social stress to be the root of their subjective sense of discomfort, our analysis of HRV leads to the conclusion that noise is more dominantly responsible for risks to health in shopping malls (Table 5). In total, levels of subjective sence of discomfort and risks for health are associated mainly with social and noise loads. While thermal load directly affects SD it is only indirectly affecting HRV by effecting human response to noise and social load. The more interesting result is that significant differences among types of places in human exposure for risk for health measured by HRV remain low when the three environmental risk factors: noise, social and thermal loads are added to the analysis. This is unlike the case of subjective sense of discomfort in which types of places remains a major factor in influencing subjective sense of discomfort independently of the effects of the three aforementioned environmental risk factors. The main limitation of the study is the fact that it is based on a limited number of measurements in only certain types of environments. There is a need to base further studies on the systematic application of Poisson distribution samples. Acknowledgments The authors thank the Israeli Science Foundation for supporting this study (ISF, grant 997/03). The authors thank Tali Hamberg, Sheli Weeg, and Shay Gendel for their assistance in the research. The authors thank all the students from Tel Aviv University and Beit Berl Academic College who took part in the field survey. Glossary CO Hz HRV LH

carbon monoxide Hertz (cycles per second) Heart Rate Variability low frequency (frequencies between 0.04 and 0.15 Hz reflecting mixed sympathetic and parasympathetic activity the last one being prevalent when breathing slowly [ms2]) HF high frequency (frequencies between 0.15 and 0.4 Hz reflect parasympathetic activity and corresponds to NeN variations (time between two heartbeats) caused by respiration: the respiratory sinus arrhythmia. Deep, even breathing activates the parasympathetic and raises the amplitude of HF [ms2]. Mental stress decreases HF activity.) LF/HF ratio (normalized units) high numbers mean dominance of sympathetic activity while low numbers mean dominance of the para-sympathetic activity. After deep and even breathing an increase reflects changes in the parasympathetic regulation. PET Physiological Equivalent Temperature SD subjective sense of discomfort

References Bitan, A., Rubin, S., 1994. Climatic Atlas of Israel for Physical and Environmental Planning and Design. Ramot Publishing Co., Tel-Aviv University, Tel-Aviv. Bjor, B., Burstrom, L., Karlsson, M., Nilsson, T., Naslund, U., Wiklund, U., 2007. Acute effects on heart rate variability when exposed to transmitted vibration and noise. Int. Archiv. Occup. Environ. Health 81, 193e199. http://dx.doi.org/ 10.1007/s00420-007-0205-0. Brown, D.K., Barton, J.L., Gladwell, V.F., 2013. Viewing nature scenes positively affects recovery of autonomic function following acute-mental stress. Environ. Sci. Technol. 47 (11), 5562e5569. http://dx.doi.org/10.1021/es305019p. Candas, V., Dufour, A., 2005. Thermal comfort: multisensory interactions? J. Physiol. Anthropol. Appl. Hum. Sci. 24, 33e36. http://dx.doi.org/10.2114/jpa.24.33. Cohen, P., Potchter, O., Matzarakis, A., 2013. Human thermal perception of coastal Mediterrenean outdoor urban environments. Appl. Geogr. 37, 1e10. http:// dx.doi.org/10.1016/j.apgeog.2012.11.001. Cohen, P., Potchter, O., Schnell, I., 2014. A methodological approach to the environmental quantitative assessment of urban parks. Appl. Geogr. 48, 87e101. http://dx.doi.org/10.1016/j.apgeog.2014.01.006. Duci, A., Chaloulakou, A., Spyrellis, A., 2003. Exposure to carbon monoxide in the Athens urban area during commuting. Sci. Total Environ. 309 (1e3), 47e58. http://dx.doi.org/10.1016/S0048-9697(03)00045-7. Epstein, Y., Heled, Y., Moran, D., Shapiro, Y., 2000. Prediction of physiological response from mathematical models. J. Israel Med. Assoc. 138 (9), 713e719. Fang, L., Wyon, D.P., Clausen, G., Fanger, P.O., 2004. Impact of indoor air temperature and humidity in an office on perceived air quality, SBS symptoms and performance. Indoor Air 14, 74e81. http://dx.doi.org/10.1111/j.16000668.2004.00276.x. Gilboa, S., 2006. The Shopping Mall as a Place in the Late Modern Age (Ph.D dissertation). Tel-Aviv University, Tel-Aviv, Israel. Gu, Z.L., Zhang, Y.W., Cheng, Y., Lee, S.C., 2011. Effect of uneven building layout on air flow and pollutant dispersion in non-uniform street canyons. Build. Environ. 46, 2657e2665. http://dx.doi.org/10.1016/j.buildenv.2011.06.028. Gullver, J., Briggs, D.J., 2003. Personal exposure to particulate air pollution in transport microenvironment. Atmos. Environ. 38 (1), 1e8. http://dx.doi.org/ 10.1016/j.atmoseznv.2003. 09.036.  Unger, J., Matzarakis, A., 2006. Assessment of the microclimatic and Guly as, A., human comfort conditions in a complex urban environment: modelling and measurements. Build. Environ. 41 (12), 1713e1722. http://dx.doi.org/10.1016/ j.buildenv.2005.07.001. Hainsworth, R., 1995. The control and physiological importance of heart rate. In: Malik, M., Camm, A.J. (Eds.), Heart Rate Variability. Futura Publishing Company, pp. 3e19. Hartig, T., Kaiser, F.G., Bowler, P.A., 2001. Psychological restoration in nature as a positive motivation for ecological behavior. Environ. Behav. 33 (4), 590e607. €ppe, P., 1999. The physiological equivalent temperature e a universal index for Ho the biometeorological assessment of the thermal environment. Int. J. Biometeorol. 43, 71e75. http://dx.doi.org/10.1007/s004840050118. €ppe, P., 2002. Different aspects of assessing indoor and outdoor thermal comfort. Ho Energy Build. 34 (6), 661e665. http://dx.doi.org/10.1016/S0378-7788(02) 00017-8. Johansson, E., Emmanuel, R., 2006. The influence of urban design on outdoor thermal comfort in the hot, humid city of Colombo, Sri Lanka. Int. J. Biometeorol. 51 (2), 119e133. Kaplan, S., 1995. The restorative benefits of nature: toward an integrative framework. J. Environ. Psychol. 15 (3), 169e182. Kaplan, R., Kaplan, S., 1989. The experience of nature: a psychological perspective. Cambridge University Press, Cambridge, U.S.A. Kaur, S., Nieuwenhuijsen, M., Colivile, R.N., 2005. Personal exposure of street canyon intersection users to PM 2.5 ultrafine particle counts and carbon monoxide in Central London, UK. Atmos. Environ. 39, 3629e3641. http:// dx.doi.org/10.1016/j.atmosenv.2005.02.046. n, M., Tyrva €inen, L., Silvennoinen, H., 2008. Determinants of Korpela, K.M., Yle restorative experiences in everyday favorite places. Health Place 14 (4), 636e652. http://dx.doi.org/10.1016/j.healthplace.2007.10.008. Kurosawa, T., Iwata, T., Dakeishi, M., Ohno, T., Tsukada, M., Murata, K., 2007. Interaction between resting pulmonary ventilation function and cardiac autonomic function assessed by heart rate variability in young adults. Biomed. Res. 28 (4), 205e211. http://dx.doi.org/10.2220/biomedres.28.205. Lin, T.P., 2009. Thermal perception, adaptation and attendance in a public square in hot and humid regions. Build. Environ. 44 (10), 2017e2026. http://dx.doi.org/ 10.1016/j.buildenv.2009.02.004. Liu, W., Lian, Z., Liu, Y., 2008. Heart rate variability at different thermal comfort levels. Eur. J. Appl. Physiol. 103, 361e366. http://dx.doi.org/10.1007/s00421008-0718-6. Matzarakis, A., Rutz, F., Mayer, H., 2007. Modelling radiation fluxes in simple and complex environments e application of the RayMan model. Int. J. Biometeorol. 51, 323e334. http://dx.doi.org/10.1007/s00484-006-0061-8. Park, B.J., Tsunetsugu, Y., Kasetani, T., Kagawa, T., Miyazaki, Y., 2010. The physiological effects of Shinrin-yoku (taking in the forest atmosphere or forest bathing): evidence from field experiments in 24 forests across Japan. Environ. Health Prev. Med. 15 (1), 18e26. http://dx.doi.org/10.1007/s12199-009-0086-9. Parsons, R., Tassinary, L.G., Ulrich, R.S., Hebl, M.R., Grossman-Alexander, M., 1998. The view from the road: implications for stress recovery and immunization.

I. Schnell et al. / Environmental Pollution 208 (2016) 58e65 J. Environ. Psychol. 18, 113e140. Pellerin, N., Candas, V., 2004. Effects of steady-state noise and temperature conditions on environmental perception and acceptability. Indoor Air 14, 129e136. http://dx.doi.org/10.1046/j.1600-0668.2003.00221.x. Peschardt, K.K., Stigsdotter, U.K., 2013. Associations between park characteristics and perceived restorativeness of small public urban green spaces. Landsc. Urban Plan. 112, 26e39. http://dx.doi.org/10.1016/j.landurbplan.2012.12.013. Potchter, O., Saaroni, H., 1998. An examination of the map of climatic regions of Israel according to the Koppen classification. Stud. Geogr. Israel 15, 179e194 (in Hebrew). Potchter, O., Oz, M., Yaacov, Y., Brenner, S., Schnell, Y., 2014. Exposure of motorbike, car and bus commuters to CO on main road in the Tel Aviv metropolitan area, Israel. Environ. Monit. Assess. 186 (12), 8413e8424. http://dx.doi.org/10.1007/ s10661-014-4013-1. Poupkou, A., Nastos, P., Melas, D., Zerefos, C., 2011. Climatology of discomfort index and air quality index in a large urban Mediterranean agglomeration. Water Air Soil Pollut. 222 (1e4), 163e183. http://dx.doi.org/10.1007/s11270-011-0814-9. Rashid, M., Zimring, C., 2008. A review of the empirical literature on the relationships between indoor environment and stress in health care and office settings problems and prospects of sharing evidence. Environ. Behav. 40 (2), 151e190. Sawasaki, N., Iwasea, S., Manoa, T., 2001. Effect of skin sympathetic response to local systemic cold exposure thermoregulatory functions in humans. Aut. Neurosci. Basic Clin. 87, 274e281. Schnell, Y., Potchter, O., Yaakov, Y., Epstein, Y., Brenner, S., Hermesh, H., 2012. Urban daily life routines and human exposure to environmental discomfort. Environ. Monit. Assess. 184, 4575e4590. http://dx.doi.org/10.1007/s10661-011-2286-1. Schnell, I., Potchter, O., Epstein, Y., Yaakov, Y., Hermesh, H., Brenner, S., Tirosh, E., 2013. The effects of exposure to environmental factors on heart rate variability: an ecological perspective. Environ. Pollut. 183, 7e13. http://dx.doi.org/10.1016/ j.envpol. 2013.02.005. Staats, H., Kieviet, A., Hartig, T., 2003. Where to recover from attentional fatigue: an expectancy-value analysis of environmental preference. J. Environ. Psychol. 23

65

(2), 147e157. http://dx.doi.org/10.1016/S0272-4944(02)00112-3. Statistical Abstract of Israel, vol. 63, 2012. Central Bureau of Statistics, 2.16. http:// www.cbs.gov.il. Sun, L., Nottrott, A., Kleissl, J., 2012. Effect of hilly urban morphology on dispersion in the urban boundary layer. Build. Environ. 48, 195e205. http://dx.doi.org/ 10.1016/j.buildenv. 2011.09.005. Toftum, J., 2002. Human response to combined indoor environment exposure. Energy Build. 34 (6), 601e606. http://dx.doi.org/10.1016/S0378-7788(02) 00010-5. €inen, L., Kagawa, T., Miyazaki, Y., 2013. PhysTsunetsugu, Y., Lee, J., Park, B.J., Tyrva iological and psychological effects of viewing urban forest landscapes assessed by multiple measurements. Landsc. Urban Plan. 113, 90e93. http://dx.doi.org/ 10.1016/j.landurbplan.2013.01.014. Ulrich, R.S., Dimberg, U., Driver, B.L., 1991a. Psychophysiological indicators of leisure benefits. In: Driver, B.L., Brown, L.R., Peterson, G.L. (Eds.), Benefits of Leisure. Venture Publishing, State College, Pennsylvania, pp. 73e89. Ulrich, R.S., Simons, R.F., Losito, B.D., Fiorito, E., Miles, M.A., Zelson, M., 1991b. Stress recovery during exposure to natural and urban environments. J. Environ. Psychol. 11, 231e248. Van den Berg, A.E., Custers, M.H.G., 2011. Gardening promotes neuroendocrine and affective restoration from stress. J. Health Psychol. 16, 3e11. http://dx.doi.org/ 10.1177/1359105310365577. VDI, 1998. Methods for the Human-biometeorological Evaluation of Climate and Air Quality for Urban and Regional Planning. Part I: Climate. VDI Guideline 3787. Part 2. Beuth, Berlin. Wang, D., Federspiel, C.C., Arens, E., 2005. Correlation between temperature satisfaction and unsolicited complaint rates in commercial buildings. Indoor Air 15, 13e18. http://dx.doi.org/10.1111/j.1600-0668.2004.00265.x. Zwack, L.M., Paciorek, C.J., Spengler, J.D., Levy, J.I., 2011. Characterizing local traffic contributions to particulate air pollution in street canyons using mobile monitoring techniques. Atmos. Environ. 45 (15), 2507e2514. http://dx.doi.org/ 10.1016/j/atmosenv. 2011.02.035.