Author: Fan, Yang; Zhiyi, Bao; Zhujun, Zhu; Jiani, Liu
Date published: April 1, 2010
Journal code: PENV
(ProQuest: ... denotes formulae omitted.)
People suffer from noise pollution every day. As one of the four major pollutions in the world, noise pollution reduces the quality of the urban environment and human health. It has been estimated that about 80 million people (approximately 20% of the population) in the European Union suffer from noise levels considered unacceptable (above 65 dB [decibels] in so-called "black areas"), while an additional 170 million people are living in "grey areas" exposed to noise levels between 55 and 65 dB (European Community, 1996). Findings from a large body of studies show that traffic noise causes non-auditory stress effects such as changes in the physiological systems (e.g., elevated blood pressure), various cognitive deficits (e.g., poor sustained attention, memory/concentration problems), sleep disturbances, psychosocial stress-related symptoms, and emotional/motivational effects (e.g., annoyance, learned helplessness) (Babisch, Beule, Schust, Kersten, & Ising, 2005; Bluhm, Berglind, Nordling, & Rosenlund, 2007; Öhrström, 2004).
Because noise pollution affects so many populations, the research on noise control is plentiful. During the 1970s and 1980s, several studies showed that plants could attenuate noise pollution by reflecting, refracting, scattering, and absorbing sound. Aylor (1972) indicated that the foliage, stems, and soil can directly attenuate noise, and that the effect of a sound barrier of vegetation on sound propagating through it was highly frequency dependent: at frequencies below about 1 kHz, the vegetation was almost transparent to noise (Bullen & Fricke, 1982; Kragh, 1981). It was conjectured that certain types of vegetation would be better at attenuating sound than others; so, factors such as the source height, microphone height, placement of the sound source, the spectrum and duration of the source, the size and density of the trees used, and the atmospheric conditions of a region had to be controlled in order to make a valid comparison between different types of vegetation (Fricke, 1984). Since the idea of sonic crystals was introduced at the end of the 20th century, the way that the arrangement and density of plants impacts noise attenuation was tested with this rising theory. Scholars discovered that a correlation existed between noise attenuation and the biological factors of plants (arrangement, density, etc.). Moreover, they found that the most effective way to create sonic crystals from trees to be used as noise-reducing devices was by planting in multiple periodic rows (Martin ez-Salaa et al., 2006). Furthermore, the visibility, tree height, belt width, and other abiological factors were important in achieving noise attenuation by foliage (Fang & Ling, 2003, 2005). All the conclusions above indicate that the research concerning the noise reduction function of foliage was focused on the noise reduction of the equivalent continuous sound level A (L^sub Acq^), the reason for the different noise attenuation at low and high frequencies, and the correlation between the foliage's biological factors and the capacity for noise reduction.
Based on the findings given by previous researchers, we presumed that the species, leaf shape, leaf tactility, and other biological factors may affects the foliage's ability to achieve noise attenuation. So we measured the noise-reducing spectrum of several plant species, compared the influences of different noise frequencies, and examined the characteristics of the foliage for noise reduction by creating a general ecotypic sound barrier made up of several plant species.
Materials and Methods
Apparatus and Materials
The electro-signal of white noise was simulated by computer and translated by an AWA6290A multichannel noise and vibration analyzer, and then inputed into a KMSEVlOlO loudspeaker as a noise source. Both data measurements and analysis use an AWA6290A multichannel noise and vibration analyzer.
Six evergreen hedges were tested in this experiment: Red Robin Photinia (Photinia fraseri), Chinese Photinia (Photinia serrulata), oleander (Nerium indicum), bamboo (Oligostachyum lubricum), deodar cedar (Cedrus deodara), and arrowwood (Viburnum odoratissimum). All of the plants are commonly used in urban green spaces in the eastern part of China and some of them are widely used in other parts of the world. For example, Red Robin Photinia and deodar cedar are commonly used in the U.S. This experiment was conducted in Zhejiang province, in the eastern part of China. All of the characteristics of the hedges are shown in Table 1.
Foliage Anangement and the Ability to Reduce Noise
We chose Red Robin Photinia (Photinia fraseri) as a sample for broad-leaved plants and deodar cedar (Cedrus deodara) for needlelikeor scalelike-leaved plants so that we might study the relationship between the arrangement of the foliage and their ability to reduce noise (Fang & Ling, 2003, 2005; MartinezSalaa et al., 2006).
The experiment was carried out at Sunshine Horticulture Nursery, which is located in Deqing county, Zhejiang province. It was sunny and breezy. The atmospheric humidity was about 30%. The ground was covered with foundation fabric. Red Robin Photinia (Photinia fraseri) and deodar cedar (Cedrus deodara) were put into individual tubs and placed in abreast and crossing arrangements. The noise source was placed 5 m in front of the hedges, while the noise receivers were placed directly in front and behind the hedges. The noise sources and receivers were 0.75 m above the ground. The noise-reducing spectrums and the LA levels of two different arrangements were recorded. Then, the density was adjusted according to the two arrangements, so as to provide 1/2 shrub breadth distance, single shrub breadth distance, and double shrub breadth distance (Figure 1). Each combination was tested five times in order to rectify the deviation in each test. Similar experiments were conducted on open ground in order to create a blank comparison control group. The differences in the mean sound pressure level resulted from the application of plants with different foliage types while the control group represented the relative attenuation value.
Characteristic of the Plants' Noise-Reducing Spectrum and the Conelation of the Biological Factors
The weather conditions were the same as above. All of the ground covers for the chosen hedges were soil, lawn, or other soft materials. The noise source was placed 5 m in front of the hedges, the noise receivers were placed closely in front and behind the hedges, and the height of the noise source and receivers were both 1.5 m above the ground. The noise-reducing spectrum (1/3 octave frequency spectrum) and the L, levels were measured five times. Tests Aeq were also conducted at sites where there were no hedges, but had similar conditions (same ground cover and similar environmental conditions). The results were subtracted from the former attenuation values in order to obtain the relative attenuation caused by the foliage. At the same time, we collected 50 leaves from each plant, and measured the average area (A) and the fresh weight of the leaves (W). The leaf tactility (T) was calculated by A/W The length and width of each leaf were also measured. The scale of the leaf shape (S) was calculated by using leaf length/leaf width.
The deodar cedar (Cedrus deodara) and Red Robin Photinia (Photinia fraseri) were planted in individual tubs. They were each around 1 m high, which was about half the height of the other four hedges. Therefore, we dropped the height of the noise source and receivers pro rata for the two hedges based on the ratio of their height to the other hedges. The width of the six hedges chosen in this experiment was between 0.8 and 3.2 m, all falling into the "noise shadow zone." The noise-reducing ability of the foliage was in a linear correlation with hedge width (Fang & Ling, 2005). We used noise average relative attenuation ... (Equation 1) as the primary guideline to compare the ability of the hedges to reduce noise. The similar heights of hedges were selected, and the experiment model was logically designed to minimize the variance caused by the height or width differences of the hedges. Then the relationship among the noise attenuation capacity of the six hedges, including leaf area (A), leaf weight (W), leaf tactility (T), and leaf shape (S), was analyzed with statistical analysis system (SAS) software.
... is the equivalent continuous Aweighted sound pressure level in front of the hedges; ... is the equivalent continuous A-weighted sound pressure level behind the hedges;
L^sub o^ is the equivalent continuous A-weighted sound pressure level of distance noise attenuation; and
W is the width of the hedges.
Results and Analysis
Correlation Between the Density, Arrangement, and ...
Table 2 shows the results of ... carried out by different arrangements and densities. The diversity of arrangement and density showed significant difference in ... (Table 3). The crossing arrangement functioned better than abreast arrangement on ... for both deodar cedar (Cedras deodara) and Red Robin Photinia (Photinia fraseri). The density was positively correlated with relative attenuation, regardless of the shape of the plants' leaves.
Arrangements' Significant Effects on the Different Frequency Segments of Noise
The third octave spectrogram of 20 Hz to 1600 Hz was achieved by the test. The result shows that the noise attenuation of diverse frequencies was significantly diverse for the two arrangements, just taking the density of 1/2 shrub breadth distance as an example (Figure 2a, 2b). The frequency spectrum was classified into a low frequency segment (20 Hz- 160 Hz), a middle frequency segment (200 Hz- 1600 Hz), and a high frequency segment (2000 Hz- 16000 Hz). The total sound pressure level of each frequency segment (L^sub p^) is defined as
where n is the number of third octave center frequencies in each segment and L^sub pi^ is the sound pressure level of each of the center frequencies (Zhang & Zhai, 2001). There is no significant difference in noise attenuation of the middle frequency segment between the two arrangements (p = .9009), which may be determined by applying ... as an independent variable, which is defined as
n = 5,
L^sub p0^' is the total sound pressure level of each frequency segment in front of the hedges,
L^sub p1^ is the total sound pressure level of each frequency segment behind the hedges, and
L^sub 0^' is the total sound pressure level of each frequency segment caused by distance.
The difference in noise frequency segment attenuation indicates that the crossing arrangement is more effective in reducing noise than the abreast arrangement. We assume that this is mainly caused by the physical laws related to the spread of sound waves. There are significant differences in the noise attenuation of high and low frequency segments. When the noise waves spread through the hedges, a lower frequency resulted in a more obvious diffraction, which resulted in lower noise attenuation values abreast arrangement than the crossing arrangement. The sound waves diffused through air and foliage, while the energy of the sound waves caused the molecules of foliage to resonate. The higher the frequency, the more intensely resonating would occur. Therefore, more energy losses result in a higher noise attenuation value. The noise presumably has to suffer more resonance in the crossing arrangement than in the abreast arrangement, resulting in different noise attenuation values of high-frequency segments.
The Plants' Noise-Reducing Spectrum
The experiment indicates that there are significant differences in ... (Table 4) and the noise-reducing spectrum for the six hedges (Figure 3). The peak value frequencies of noise-reducing ability are determined by the foliage species. Even if there were several hedges with partial same peak value frequencies, the degrees of noise-reducing ability were different. For example, the peaks for Chinese Photinia (Photinia serrulata) occur at the frequencies of 80 Hz, 160 Hz, 400 Hz, 1000 Hz, 2000 Hz, and so on; and there are peaks at 250 Hz and 800 Hz for bamboo (Oligostachyum lubricum); 1000 Hz is one of the peak values for arrowwood (Viburnum odoratissimum), oleander (Nerium indicum) and Chinese Photinia (Photinia serrulata); but the degrees of noise-reducing ability of these plants are different. Therefore, it is unscientific to ignore the differences among the different plants, while considering the plants independent noise-reducing materials. In other words, every species has its own noise attenuation spectrum according to the frequencies.
By analyzing the noise-reducing spectrum of six hedges, we find that there are some commonalities in the noise-reducing spectrum of the five evergreen, broad-leaved hedges except for deodar cedar (Cedrus deodara). The noise attenuations of low frequency caused by the five plants are not significant. The standard deviations are large and the curves are not clearly separate, so the diversity is not significant among the five hedges. The noise attenuations of middle frequency fluctuated considerably. That is, fluctuations in peak-valley measurements were apparent so when the standard deviations decreased by degrees the curves started to separate. The noise attenuations increased significantly in the high frequency segment, and as the standard deviations decreased, the curves notice ably separated. All in all, the results indicate that plants with different morphological and physiological characteristics (evergreen, broad-leaved plants) have different effects in reducing noise, especially the noise on the middle and high frequency segments.
The noise-reducing capability of the plants is associated with leaf shape. The noise attenuation value for plants with long and narrow leaves, such as oleander (Nerium indicum) and bamboo (Oligostachyum lubricum) are lower than the plants with ovate or elliptic types of leaves such as arrowwood (Viburnum odoratissimum) and Chinese Photinia (Photinia serrulata). The particular noise-reducing spectrum for deodar cedar (Cedrus deodara), which is the only conifer of the six hedges, shows that the morphological and physiological characteristics of plants remarkably affect the noise attenuation. Consequently, we presume that the noise-reducing ability is closely related to the leaf shape, leaf tactility, and other morphological and physiological characteristics. And the findings mentioned above imply that the species diversity of the noise-reducing spectrum may be a potential factor that can buffer against certain frequencies. For example, the deodar cedar would be good at attenuating noises created by traffic, which is mainly made up of middle and low frequencies, because of its effective low-frequency noise-reducing spectrum.
Multiple Regression Model of ...
The model was established by using leaf weight, tactility, and shape as independent variables, and ... as a dependent variable (in this model p < .001). Leaf area is ignored in this model since when it is involved as an independent variable this model is not statistically significant (p = .1128). The model is defined as
W is the leaf weight,
T is the leaf tactility, and
S is leaf shape.
The leaf shape and leaf tactility are negatively correlated, while leaf weight is positively correlated with the average relative attenuation. It indicates that leaves that have a long and narrow shape (a larger length and width ratio) have a smaller area for acoustic reflection or acoustic refraction than the ovate or elliptic shaped leaves. Hence they have a smaller noise attenuation value. The thick leaves are always narrow and long or spread, so the plants can centralize their biomass, and the visibility of the hedge is also large. As a result, the noise attenuation is lower (Fang & Ling, 2003). The leaf tactilities of arrowwood (Viburnum odoratissimum), Red Robin Photinia (Photinia/raseri), and Chinese Photinia (Photinia serrulata) art positively correlated to their average relative attenuation values. This indicates that when leaf shape remains unchanged, the leaf weight may affect the noise attenuation value much more than leaf tactility. The higher the leaf tactility is, the more the leaf weighs, the larger the biomass of the plant, and the more energy the plant will lose when the sound wave spreads through the hedge. As a result, the noise attenuations are significantly increased.
Plants reduce noise through reflection, refraction, absorption, interference, and diffraction. Special arrangement can significantly increase the noise attenuations for certain frequencies. The results show that compared to abreast arrangements, the crossing arrangement has more significant effect towards reducing noise for middle and high frequencies. The reason for this phenomenon can be referred to the rule of sound interference and diffraction. The phenomenon also agrees with Martinez-Salaa and co-authors (2006). They demonstrated that some periodic tree configurations work like sonic crystals, and that the existence of attenuation peaks and their position within certain frequencies seem to be an exclusive function of the parameters that govern the behavior of sonic crystals and are not related to other factors such as the type of land, foliage, etc. The result of our experiment, however, shows that the attenuation peak frequencies closely correlate with the plant species. The different arrangements determine the degree of noise attenuation. For example, the two arrangements of Red Robin Photinia (Photinia fraseri) both result in attenuation peaks at 2000 Hz, but the attenuation value of the crossing arrangement is higher than the abreast arrangement.
Density, height, length, and width of the green sound barriers are the most important factors to consider in reducing noise (Cook & Haverbeke, 1974). Shrubs are most effective in reducing noise due to their ability to scatter noise as a result of their dense foliage and branches (Fang & Ling, 2003). Aylor (1972) pointed out that foliage reduces sound transmission substantially, especially at the higher frequencies where scattering is enhanced. The effectiveness of the foliage increases with the increase of leaf width and leaf weight, which is proven by the multiple regression model of average relative attenuation in our study The result shows that the noise-reducing spectrums of different plants are significantly different. The noise attenuations caused by hedges are mainly due to the foliage, stems, leaf shape, leaf tactility, and other biological factors. For the foliages with diverse morphological and physiological characteristics, the sound wave is a kind of multiple energy wave; the attenuation of sound waves mainly depends on the species. Our study makes an additional contribution to the research on noise and health and the influence of natural environments by showing that the noise attenuation of green sound barriers can be enhanced by assembling diverse plants whose noise-reducing spectrums are complementary. Since those factors are determined by the species, it is important to consider the multiple regression model, which considers leaf characteristics such as shape, weight, and tactility to establish the noise attenuation comparisons based on the type of foliage use. This is just a preliminary study on the noise-reducing spectrum of plants, however. Some limitations in our study need to be acknowledged. First, the number of species of plants used in the research is relatively small, especially regarding the coniferous type of hedges. The impact of leaf shape on the noise-reducing spectrum needs to be verified by other conifers after the characteristic of low-frequency reduction demonstrated by the deodar cedar (Cedrus deodara) was discovered. Second, further research needs to examine the combination of diverse plants, such as a green sound barrier made up of a row of Red Robin Photinia arranged in front of a row of deodar cedar. Thus, well-designed longitudinal studies can offer valuable knowledge of potential causeeffect relationships and health benefits of green sound barriers in noise-polluted environments.
The ultimate goal of this research is to design the most effective ecotypic sound barrier with plants. By reviewing previous studies, and through our own research, some general suggestions of designing ecotypic sound barrier are concluded. (1) Choose the plants with a noisereducing spectrum similar to the environmental noise spectrum, or combine several plants in order to achieve the most effective noise attenuation. Deodar cedar (Cedrus deodara), for example, reduces low frequency noise more effectively than the other five plants studied, even though its ... is negative. It is possible to combine deodar cedar (Cedrus deodara) with other plants in order to take advantages of environment noise control properties. The biological factors of the plants such as leaf shape, leaf weight, and leaf tactility directly affect the noise attenuation value, including ... and the frequencies. Each plant has its own noisereducing spectrum. Noise attenuation will be improved by combining plants with reciprocal noise-reducing spectrums. (2) Use crossing arrangements to reduce low-frequency noise. This arrangement is much more effective than the abreast arrangement. The attenuation peaks are determined by the species, while the arrangements determine the attenuation degrees. (3) Fulfill the basal demand for the length, width, and height of hedges. The hedgerow is effective in noise reduction when it is high and wide if the distance between the noise source and receiver is less than eight times the tree height (Fang & Ling, 2005).
Acknowledgements: The authors wish to thank the anonymous reviewers for their suggestions and remarks concerning the manuscript, and express deep gratitude to Mr. Yi Zhu for his embellishment and Mr. Weiqiang Li and Mr. Zizheng Hu for their exceptional support of this research.
Aylor, D. (1972). Noise reduction by vegetation and ground. The Journal of the Acoustical Society of America, 51(1B), 197-205.
Babisch, W., Beule, B., Schust, M., Kersten, N., & Ising, H. (2005). Traffic noise and risk of myocardial infarction. Epidemiology, 1 6(1), 33-40.
Bluhm, C, Berglind, N., Nordling, E., & Rosenlund, M. (2007). Road traffic noise and hypertension. Occupational & Environmental Medicine, 64(2), 122-126.
Bullen, R., & Fricke, F (1982). Sound propagation through vegetation. Journal of Sound and Vibration, 80(1), 11-23.
Cook, D.I., & Haverbeke, D. EV (1974). Trees and shrubs for noise abatement (Bulletin RB246). Lincoln: University of Nebraska College of Agriculture Experiment Station.
European Community. (1996). Commission green paper on future noise policy. Retrieved August 16, 2007, from http://ta.europa.eu/environment/noise/greenpap.htm
Fang, C.E., & Ling, D.L. (2003). Investigation of the noise reduction provided by tree belts. Landscape and Urban Planning, 63(4), 187-195.
Fang, C.E., & Ling, D.L. (2005). Guidance for noise reduction provided by tree belts. Landscape and Urban Planning, 71 (1), 29-34.
Fricke, E (1984). Sound attenuation in forest. Journal of Sound and Vibration, 92(1), 149-158.
Kragh, J. (1981). Road traffic noise attenuation by belts of trees. Journal of Sound and Vibration, 74(2), 235-241.
Martinez-Salaa, R., Rubio, C., García-Raffi, L.M., Sánchez-Pérez, J.V., Sánchez-Pérez, E.A., & Llinares, J. (2006). Control of noise by trees arranged like sonic crystals. Journal of Sound and Vibration, 291(1-2), 100-106.
Öhrström, E. (2004). Longitudinal surveys on effects of changes in road traffic noise annoyance, activity disturbances, and psychosocial well being. The Journal of the Acoustical Society of America, 115(2), 719-729.
Zhang, B.J., & Zhai, G.Q. (2001). The research of environmental noise. Hangzhou, China: Zhejiang University Press.
Bao Zhiyi, PhD
Zhu Zhujun, PhD
Liu Jiani, MSc
Corresponding Author: Zhiyi Bao, Professor, Vice Dean of the School of Landscape Architecture, Zhejiang Forestry University, North Huancheng Road 88, Linan, Hangzhou, China. E-mail: firstname.lastname@example.org.