Noise Reading Types Continuous Intermittent Impulse

Intermittent Noise

Intermittent noise: Noise that stops and starts, usually at irregular intervals, is considered to be an intermittent noise.

From: Environmental Noise Pollution , 2014

Volume 5

A. Muzet , in Encyclopedia of Environmental Health (Second Edition), 2011

Reduced Sleep Amount

The total duration of sleep can be reduced by lengthening of the falling asleep process, nocturnal awakenings, or an earlier final awakening. It has been shown that intermittent noises with a maximal intensity of 45  dB(A) can delay sleep onset by a few minutes to 20   min, whereas a level of 55   dB(A) could represent an approximate average value for the awakening threshold during nocturnal sleep. Intrasleep awakenings occur in any sleep stage with variable threshold, but for a given sleep stage this threshold decreases with increasing time spent asleep. This is why late night hours are more easily disturbed by ambient noises. Therefore, premature early morning awakening increases the risk of sleep debt.

The awakening threshold is much higher in stages 3 and 4 (slow wave sleep) than in stages 1 and 2. The threshold observed in REM sleep is variable and depends mainly on the noise significance. In this particular sleep stage, the sleeper will react more easily to significant noise such as the call of his or her own name or to a familiar noise. Similarly, an alarm sound will be more effective than a sound with similar intensity but with a neutral significance. This shows that the reactivity of a sleeper is not always proportional to the noise intensity and that the significance of the noise is sometimes more important than its physical characteristics. Sensitivity to noise depends also on the age of the sleeper. Elderly subjects in particular seem to be more reactive to the nocturnal noises. A confounding variable is the fact that melatonin secretion decreases in adults after the age of 55 (also more importantly less slow wave sleep (stages 3+4) and more stage 1), the effect of which is to produce a greater amount of sleep fragmentation. This natural fragmentation of their nocturnal sleep explains why their spontaneous awakenings are also more frequent than in younger adults. Therefore, while awaken they hear the ambient noises and they often consider these events as responsible for their awakenings.

It is difficult to find a clear difference due to gender, although in the general attitude toward ambient noise, women seem to be more reactive than men. In fact, they complain more but objectively male sleep is generally more disturbed than female. Finally, it must be stressed that in the past decades, more and more people have been concerned with unusual work schedules. Some often work at night and sleep during the daytime when ambient noise level is generally much higher. This could explain why these workers complain very much about their sleep being disturbed and of poor quality. Also, there is desynchronosis where the habitual entrainment of circadian rhythms is lost.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780444639516002497

Principles of Environmental Noise

Enda Murphy , Eoin A. King , in Environmental Noise Pollution, 2014

There are many adverse effects associated with exposure to environmental noise. These can range from hearing impairment to sleep disturbance to annoyance and even cardiovascular disorders. These relationships are explored in more detail in Chapter 3 . In the case of environmental noise, annoyance refers to the non-specific disturbance from noise and may include the reduced enjoyment of an outdoor space or the necessity of keeping one's windows shut at home as a result of noise immission. The level of annoyance an individual experiences due to noise is a complex issue and is governed by numerous and (often) subjective factors. Intermittent noise, noise that stops and starts, is considered to be more annoying than continuous noise while the presence of audible tones (one frequency being heard above other frequencies, e.g., a high-pitched whine) also increases annoyance. Environmental noise also tends to be more bothersome during summer than winter and research suggests that marital status and gender may also play a part a role in the feeling of annoyance caused by noise exposure ( Abo-Qudais and Abu-Qdais, 2005; Miedema et al., 2005).

Box 2.1

The Study of Acoustics

Acoustics is the study of the doctrine of sounds. In 1964 Robert B. Lindsay described the scope of acoustics in the broad fields of Earth Sciences, Engineering, Life Sciences and the Arts. He developed a 'Wheel of Acoustics' to describe how acoustics relates to these fields. This highlighted succinctly the inter-disciplinary nature of the study of acoustics.

The word 'acoustics' is believed to have been introduced to the English language by Archbishop Narcissus Marsh (1638–1713). Archbishop Marsh served as provost of Trinity College Dublin (1679–1683) and was responsible for building the first public library in Ireland in 1701. He also invented the word 'microphone' – almost 200 years before the device was invented (An Introductory Essay to the Doctrine of Sounds, 1683).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124115958000021

Transportation Noise

Enda Murphy , Eoin A. King , in Environmental Noise Pollution, 2014

5.4.1 Road Traffic Noise

Current road traffic noise prediction methods are outdated and are being used in situations for which they were never originally intended. CNOSSOS-EU represents a significant step forward in this regard. It will incorporate many aspects of today's best practices in noise emission and sound propagation modelling. In terms of frequency analyses, it is proposed that CNOSSOS-EU will perform calculations across octave bands. This is consistent with the recommended interim method for road traffic (although two extra octave bands outside the scope of the recommended interim method, at centre frequencies of 63 and 8000   Hz, are considered in CNOSSOS-EU) and certainly represents an improvement when compared to methods that only predict an overall A-weighted sound pressure level. However, for detailed assessments involving annoyance or tonal assessments – which may be needed, for example, with the increasing number of electric vehicles on major roads – a detailed consideration of frequency spectra for different vehicle types is required and the CNOSSOS-EU method will have to be adapted to perform such studies.

The manner in which road traffic noise is divided into vehicle categories is an aspect that will be improved by CNOSSOS-EU. The current default approximation assumes just two categories (light and heavy), whereas CNOSSOS-EU divides vehicles into five classes in accordance with definitions set out in Directive 2007/46/EC. However, it is worth noting that the Harmonoise model proposed five broad vehicle categories which were divided into 18 subcategories (Jonasson et al., 2004). The intention was to model the five main categories initially but, as new data were collected for each subcategory, it would then be possible to model each subcategory. At present, determining datasets for 18 separate vehicle categories is probably beyond the capabilities of most noise-mapping authorities but such detailed data may exist in the future. Today, some authorities may even struggle with the proposed five categories. Some of the vehicle categories set out in Directive 2007/46/EC are classified according to weight. This may be troublesome for authorities who do not have the capability of capturing vehicle weight with existing traffic counters.

The treatment of low-noise road surfaces is an area in need of further research. The variation in acoustic properties of road surfaces is large, and there is no common procedure for the assessment of the acoustic properties of road surfaces (Kephalopoulos et al., 2012). The CNOSSOS-EU method will allow Member States apply their own regional road surface corrections, provided these corrections are documented and reported. Ideally, corrections for low-noise road surfaces should be derived from national datasets to account for national differences. These corrections should all be compared to the hypothetical reference surface described in CNOSSOS-EU and documented. This may eventually lead to a European road surface database and may facilitate the development of more effective low-noise road surfaces.

Most road traffic noise prediction methods in use today mix engine noise and rolling noise because emission quantities were originally derived from single microphone pass-by measurements. The CNOSSOS-EU method of separating rolling noise and propulsion noise is a welcome development and is now considered best practice internationally. However, in order to maximise the effectiveness of this development, the model should be refined to include separate source heights, as initially proposed in the Harmonoise method. This would allow the contribution of each source mechanism to be divided between multiple source positions. If both source mechanisms are combined at one position (usually close to the ground), the contribution of rolling noise and engine noise cannot be separated and the effectiveness of some mitigation measures might be either over- or underestimated. For example, a noise barrier beside a major road might not be designed sufficiently high if the engine noise from a heavy vehicle is modelled at a height of 0.05   m (see Figure 5.10). This is only likely to be an issue at specific locations where barriers and receivers are close to the road, so might it not be enough to warrant the related increase in computational time, but the model should be capable of performing more detailed calculations when desired. This would enable the improved assessment of potential mitigation measures.

Figure 5.10. Sketch of different source positions at the influence on noise mitigation measures.

CNOSSOS-EU also includes corrections for the acceleration and deceleration of vehicles. These corrections are important because the acoustic characteristics of intermittent traffic flow are considerably different to free-flowing traffic in free-field conditions. Yet for the purpose of strategic noise maps, the effects of acceleration and deceleration can be neglected (Kephalopoulos et al., 2012) because, generally speaking, the average sound pressure level for accelerating and decelerating traffic does not depart significantly from the level assumed for a steady speed across a junction (Watts, 2005). However, this is very much associated with the use of energy based indicators such as L_den and L_night . Academic research has pointed out that current noise measurement techniques and noise indicators do not readily accommodate the assessment of intermittent noise of large vehicles driving at night which is associated with high levels of community annoyance ( Schreurs et al., 2011). Accordingly, annoyance assessments should account for varying noise at junctions and this requires alternative indicators. The form that these indicators ultimately take will dictate how the various emission models should be developed.

Box 5.11

Supplemental Indicators to Calculate Annoyance

In the first phase of noise mapping, supplementary noise indicators, see Chapter 4, were rare and confined to indicators such as L_max or L_eq at 2   m. Many different noise indicators exist and their use may maximise the value of strategic noise mapping. They are often used to take account of situations that are not appropriately described with the recommended EU noise indicators, L_den and L_night . Examples include, L_max , Perceived Noise Level, Sound Exposure Level (SEL), or even % Highly Annoyed (%HA) and % Highly Sleep Disturbed (%HSD). It is also worth questioning if the dose response relationships describing %A or %HA (which were based on extensive surveys carried out in USA and Northern Europe) are also applicable to the polar and subtropical climates of Northern and Southern EU Member States, respectively (Wolde, 2003).

Finally, another limitation of noise prediction methods lies in their inability to include driver behaviour and how it varies from one nation to the next in calculations. For example, different attitudes to horn use in Brazil and England have been cited as a reason for the varying levels of accuracy of the CRTN method (the UK's road traffic noise prediction method) when utilised in the two countries (Filho et al., 2004). Of course, given that no method claims to predict horn use or driver behaviour, it is probably somewhat harsh to label this as a limitation of calculation methods per se. Some standards do go beyond what would normally be considered within the scope of a prediction model. The German RLS 90 method, for example, includes a method for calculating noise for parking lots which is uncommon for most calculation methods (Steele, 2001). It may be appropriate for future versions of CNOSSOS-EU to consider aspects outside the scope of the current model as further research is conducted in the area.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124115958000057

Natural Time Analysis of Seismic Time Series

Nicholas V. Sarlis , ... Panayiotis A. Varotsos , in Complexity of Seismic Time Series, 2018

7.2.3 The Two Origins of Self-Similarity and Their Distinction in the Case of Seismicity

A large variety of natural systems exhibit irregular and complex behaviours, which at first look seem to be erratic, but in fact possess scale-invariant structure, e.g., see Peng et al. (1995a) and Kalisky et al. (2005). A stochastic process $X(t)$ is called self-similar with index $H>0$ if it has the property (Lamperti, 1962)

(7.4) $X(\lambda t)\underline{\underline{d}}{\lambda }^{H}X(t)\forall \lambda >0.$

where the equality concerns the finite-dimensional distributions of the process $X(t)$ on the right- and the left-hand side of the equation (not the values of the process).

In several systems, this nontrivial structure points to long-range temporal correlations which alternatively means that self-similarity results from process memory only (but we stress that long-range temporal correlations do not automatically imply self-similarity of a process, e.g., Varotsos et al. (2006b)). This is the case, e.g., of fractional Brownian motion (fBm) or of SES activities. Alternatively, the self-similarity may solely result from the process increments infinite variance (heavy tails in the distribution). This is the case, e.g., with Levy stable motion. Note that Levy stable distributions, which are followed by many natural processes (e.g., see Tsallis et al., 1995, 1996), have heavy tails and their variance is infinite (Scafetta and West, 2005; Weron et al., 2005; Ausloos and Lambiotte, 2006). In general, the distinction of these two origins of self-similarity, i.e., process memory and process increments infinite variance, which may coexist, is a difficult task. This has been attempted in Varotsos et al. (2006b) by employing natural time analysis and further investigated in Sarlis et al. (2009b).

According to Varotsos et al. (2006b), the use of natural time analysis can lead to the identification of the origin of self-similarity as follows: First, if self-similarity results from the process memory only, the ${\kappa }_1$ value should change to ${\kappa }_u=1/12$ for the (randomly) shuffled data. Second, if the self-similarity exclusively results from process increments 'infinite' variance, the ${\kappa }_{\mathrm{1,}p}$ value, at which the probability distribution function (PDF) $P({\kappa }_1)$ (see Section 7.3.1) maximizes, should be the same (but different from ${\kappa }_u$ ) for the original and the randomly shuffled data. This procedure answers, e.g., to the fundamental problem of distinguishing between stochastic models characterized by different statistics, e.g., between fractal Gaussian intermittent noise and Levy-walk intermittent noise, that may equally well reproduce some patterns of a time series ( Scafetta and West, 2004, 2005). When both sources of self-similarity are present in the time series, quantitative conclusions on their relative strength can be obtained on the basis of Eqs. (7.12) and (7.13) of Varotsos et al. (2006b) that will now be explained.

These two equations relate either the expectation value $\mathrm{E}({\kappa }_1)$ of ${\kappa }_1$ in the actually observed time series, or the expectation value $\mathrm{E}({\kappa }_{\mathrm{1,}shuf})$ in a randomly shuffled time series, when a (natural) time window of length $l$ is sliding through the time series $Q_k\ge 0,k=1,2,\dots N$ . For such a window, starting at $k=k_0$ , the quantities $p_j=Q_{k_0+j-1}/{\sum }_{m=1}^l{Q}_{k_0+m-1}$ in natural time are defined and Varotsos et al. (2006b) find that $\mathrm{E}({\kappa }_1)$ in the actually observed time series is given by

(7.5) $\mathrm{E}({\kappa }_1)={\kappa }_{1,\mathbb{M}}+\sum _{\mathrm{all\; pairs}}\frac{{(j-m)}^2}{l^2}\mathrm{Cov}(p_j,p_m),$

where ${\kappa }_{\mathrm{1,}\mathbb{M}}$ is the value of ${\kappa }_1$ corresponding to the time series of the averages ${\mu }_j\equiv \mathrm{E}(p_j)$ of $p_j$ , i.e.,

(7.6) ${\kappa }_{\mathrm{1,}\mathbb{M}}=\sum _{j\mathrm{=1}}^l{(j/l)}^2{\mu }_j-{\left(\sum _{j\mathrm{=1}}^l{\mu }_{j}j/l\right)}^2,$

and $\mathrm{C}ov(p_j,p_m)$ stands for the covariance of $p_j$ and $p_m$ defined as

(7.7) $\mathrm{C}ov(p_j,p_m)\equiv \mathrm{E}[(p_j-{\mu }_j)(p_m-{\mu }_m)],$

while the variance of $p_j$ is given by

(7.8) $\mathrm{Var}(p_j)=\mathrm{E}\left[{\left(p_j-{\mu }_j\right)}^2\right].$

The symbol ${\sum }_{\mathrm{all}\mathrm{pairs}}$ stands for ${\sum }_{j\mathrm{=1}}^{l-1}{\sum }_{m=j+1}^l$ . Eq. (7.5) reveals that $\mathrm{E}({\kappa }_1)$ is determined by two factors that involve: (1) the correlation of the data as reflected in the averages ${\mu }_j$ , e.g., due to a decrease in the magnitude of aftershocks in an earthquake time series, and (2) the covariance term which sums up the correlations between all natural time lags up to $l-1$ .

On the other hand, $\mathrm{E}({\kappa }_{\mathrm{1,}shuf})$ obtained upon randomly shuffling the original time series is found (Varotsos et al., 2006b) to be

(7.9) $\mathrm{E}({\kappa }_{\mathrm{1,}shuf})={\kappa }_u\left(1-\frac{1}{l^2}\right)-{\kappa }_u(l+\mathrm{1)}\mathrm{Var}(p)$

(note that for the shuffled data $\mathrm{Var}(p_j)$ is independent of $j$ , and hence we merely write $\mathrm{Var}(p)\equiv \mathrm{Var}(p_j)$ ). If $Q_k$ do not exhibit heavy tails and have finite variance, Eq. (7.9) rapidly converges to $\mathrm{E}({\kappa }_{\mathrm{1,}shuf})={\kappa }_u$ . Otherwise, $\mathrm{E}({\kappa }_{\mathrm{1,}shuf})$ differs from ${\kappa }_u$ , and the difference

(7.10) $\mathrm{\Delta }E({\kappa }_{1,shuf})\equiv {\kappa }_u\left(1-\frac{1}{l^2}\right)-\mathrm{E}({\kappa }_{1,shuf})={\kappa }_u(l+1)\mathrm{Var}(p)$

provides a measure of the process increments 'infinite' variance. Comparing the results obtained from Eqs. (7.5), (7.9) and (7.10), we can draw quantitative conclusions on the existence of temporal correlations in a real time series even if the process increments exhibit 'infinite' variance. This procedure was followed by Sarlis et al. (2009b) demonstrating the existence of temporal correlations between earthquake magnitudes in the seismicity of Southern California. The same method was later employed (Sarlis, 2011; Sarlis and Christopoulos, 2012) for the identification of the correlations between successive magnitudes in global seismicity.

Calculating the ${\kappa }_1$ value by means of a window $l\mathrm{=6}$ to 40 consecutive events sliding through either the original earthquake catalogue or a shuffled one (see Section 7.3.1), Varotsos et al. (2006b) obtained the following results for the SCEC (Southern California) as well as for the JMA earthquake catalogue (Japan): In both catalogues, the most probable values of ${\kappa }_1$ , denoted by ${\kappa }_{\mathrm{1,}p}$ , are found to be ${\kappa }_{\mathrm{1,}p}\approx 0.066$ for the original data, while they are ${\kappa }_{\mathrm{1,}p}\approx 0.064$ for the surrogate data. Both these ${\kappa }_{\mathrm{1,}p}$ values differ markedly from the value ${\kappa }_u=1/12$ of the 'uniform' distribution. This could be interpreted as reflecting that the self-similarity mainly originates from the process increments infinite variance. In addition, since the ${\kappa }_{\mathrm{1,}p}$ value of the original earthquake data does not differ considerably from the value ${\kappa }_1\approx 0.070$ identified (Varotsos et al., 2002a, 2003a,b, 2011c,d) in infinitely ranged temporal correlations, this indicates the importance of temporal correlations rather than their absence in the earthquake catalogues. In other words, the temporal correlations between earthquake magnitudes are responsible for the difference between the value of ${\kappa }_{\mathrm{1,}p}\approx 0.064$ of the surrogate data from the value of ${\kappa }_{\mathrm{1,}p}\approx 0.066$ of the original data (both these values have a plausible uncertainty of ±0.001; Sarlis et al., 2009b). At this point, it is worthwhile mentioning that by means of natural time analysis of a generalized Cantor set (multiplicative cascade), a theoretical interrelation between ${\kappa }_{\mathrm{1,}p}$ of the (randomly) shuffled earthquake data and the parameter $b$ of the Gutenberg–Richter law can be found (Sarlis et al., 2009b)

(7.11) ${\kappa }_{1,p}=\frac{2^{\frac{3}{2b}}}{3{\left(1+2^{\frac{3}{2b}}\right)}^2}.$

Recall that according to Gutenberg and Richter (1954) the (cumulative) number of earthquakes with magnitude greater than (or equal to) $M,N(\ge M)$ , occurring in a specified area and time, is given by

(7.12) $N(\ge M)={10}^{a-bM},$

where the constant $a$ gives the logarithm of the number of earthquakes with magnitude greater than zero (e.g., Shcherbakov et al., 2004). Eq. (7.11), if we just adopt a reasonable value of $b$ , i.e., $b\approx 1$ , leads to a value of ${\kappa }_{\mathrm{1,}p}$ that is very close to 0.064, thus agreeing with the aforementioned ${\kappa }_{\mathrm{1,}p}$ value deduced by Varotsos et al. (2006b) from the natural time analysis of the shuffled experimental data of SCEC and Japan.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128131381000079

A review of the environmental parameters necessary for an optimal sleep environment

Zachary A. Caddick , ... Erin E. Flynn-Evans , in Building and Environment, 2018

3 Noise

Exposure to noise can disrupt sleep quality and quantity [61]. The magnitude of sleep disruption conferred by noise depends on the decibel level (dB), the frequency and pitch, duration (i.e. continuous, intermittent, or impulsive) and whether the noise is meaningful (e.g. a familiar voice). A World Health Organization working group report on noise determined that there is a causal relationship between nighttime noise exposure and self-reported sleep disturbances, use of pharmaceuticals, self-reported health problems, and insomnia-like symptoms [122]. The same group issued guidelines for noise exposure during sleep, setting the limit for average level between 30 and 40 dBA, citing that higher levels of noise in the sleep environment leads to changes in the duration of sleep stages and increase sleep fragmentation.

Table 2 shows the literature describing the association between noise exposure and sleep disruption. There have been several high quality randomized (category 1A) or quasi-experimental (category 2A) studies using EEG that confirm that the auditory arousal threshold that causes a transition from sleep to wake varies between individuals and sleep stage [70,87]. Awakening due to noise exposure <50 dBA is more likely in shallow stages of sleep (i.e. stages 1 & 2), where louder noises [42,43,69,121] or noises that are in the low frequency range (∼500 Hz [15], are required to cause waking from deeper stages of sleep (i.e. stages 3 & 4. It is notable that the arousal threshold in REM sleep is not easily determined due to the influence of dreaming (described in Ref. [87]). Complete awakening from sleep appears to be dependent on the frequency of repetition of the noise in addition to the volume, with more frequent pulses causing more sleep disruption [70]. Similarly, the auditory arousal threshold has been shown to change with repeated exposure to noise [11] and others have shown that there are large inter-individual differences in noise sensitivity [62]. The awareness that sleeping individuals have of their surroundings can also contribute to awakenings, such as speaking a sleeper's name [78], the sound of human voices [49,50], and household activity [95].

Table 2. Summary of studies describing noise levels and their influence on sleep.

Noise Source	Type of Disruption	Sleep Measure	Noise Level	Type of Noise	Source
Church bells	Dose response between awakening and noise level	EEG questionnaire	30-70 dB	Intermittent	[14]
Rail noise	Higher noise level: increased arousals, awakening, S1, SWS, sleep stage transitions, reduced TST	EEG	30-65 dBA	Intermittent	[6,26,94]
Aircraft noise	Higher modeled aircraft noise = no difference in sleep insufficiency compared to lower modeled noise	Questionnaire	Modeled noise of <60, 60–65, >65 dBA	Intermittent	[34]
Aircraft noise	Reduced sleep quality	Actigraphy self-report	Quiet control, 60 dBA	Intermittent	[96]
Aircraft noise	Higher noise level: increased awakening, arousals sleep-stage transitions, reduced SWS, TST	EEG	30-65 dBA	Intermittent	[5–7]
Traffic noise	Increased latency, awakening, consumption of sleeping pills, reduced sleep quality	Questionnaire self-report	>45–75 dBA	Not reported	[103,114]
Traffic noise	Reduced TST, increased sleep latency, awakening	Actigraphy sleep logs	40-45 dB (inside bedroom)	Not reported	[81]
Traffic noise	Higher noise level: increased awakening, S1, arousals sleep-stage transitions, reduced SWS, TST, WASO	EEG	30-60 dBA	Intermittent	[41,83,107,120]
Traffic and rail noise	Individuals living in a dwelling with a quiet side reported reduced sleep quality	Questionnaire	Modeled noise of <40,40–60, and >60 dBA	Intermittent	[10]
Aircraft, traffic, and rail noise	Increased sleep latency, WASO, S1, reduced TST, sleep efficiency, SWS, REM	EEG	32 dB vs. 39, 44, 50 dBA	Continuous	[60]
Wind turbine noise	Dose response relationship with poorer sleep associated with nearness to wind turbines	PSQI, ESS	40-52 dBA	Continuous	[71]
Wind turbine (pre/post installation)	No differences in EEG sleep, reduced self-reported sleep quality	EEG self-report	37 dBA (pre), 37 dBA (post)	Continuous	[37]
Wind turbine noise	No association between sleep outcomes and level of noise	Actigraphy, PSQI self-report	<25, 25–30, 30–35, 35–40, 40–46 dBA	Continuous	[63]
General noise in the sleep environment	Fragmentation of sleep and impact on duration of various sleep stages	N/A	35 dBLAmax, inside	Not specified	World Health Organization Limit [122]

dBA = A-weighted decibels (where low frequencies are reduced); dBL_{Amax, inside} = maximum levels per event inside a bedroom; dB = decibels; EEG = electroencephalogram; PSQI = Pittsburgh Sleep Quality Index; ESS = Epworth Sleepiness Scale.

There have been few category 1 or EEG studies that have been conducted evaluating the impact of continuous noise on sleep. The two experimental EEG studies (category 1A) that have been conducted suggest that exposure to continuous ambient noise of 39 dBA or greater is associated with increased night waking, shorter sleep duration, and poorer sleep quality, including reduced REM sleep [60,97 ]. In contrast, continuous noise of 62 dB containing a blend of 1–22.05 kHz has been shown to protect sleep by masking the influence of exposure to other intermittent noises [ 101]. In the natural environment, continuous noise pollution typically emanates from sources such as exposure to wind turbines, however, it is unclear whether exposure to wind turbine noise influences sleep. One study found a dose response relationship with poorer sleep being associated with proximity to wind turbines generating 40–52 dBA using the Pittsburgh Sleep Quality Index (PSQI) and Eppworth Sleepiness Scale (ESS; category 3C) [71]. However, another study found no association between sleep outcomes and level of wind turbine noise ranging from <25 dBA to 46 dBA using PSQI and actigraphy (category 3B) [63]. Similarly, a study that compared EEG-measured sleep before and after installation of wind turbines in a neighborhood found no differences in objective or subjective sleep outcomes (category 1A), however, the noise exposure in that study was the same in both conditions (37 dBA, [37]. These findings suggest that quiet environments are better for sleep, but that continuous noise may be useful in dampening intermittent noises. More research is needed to determine what level and frequency of continuous noise is detrimental to sleep outcomes.

Intermittent noise is generally perceived to be more disruptive to sleep than continuous noise. There are numerous sources of intermittent noise that can infiltrate the bedroom environment. Several randomized studies using EEG in controlled environments (category 1A) have shown that exposure to intermittent noises such as simulated door slamming [115], passing trains [94] [6], aircraft flyovers [5] [6], and traffic noise [6,83] are associated with increased night waking, increased arousals from sleep, increased slow wave sleep, increased sleep stage transitions, and reduced total sleep time. Studies in the field largely confirm these findings. Several high quality observational studies using EEG have shown that intermittent exposure to noise in the environment is associated with sleep disruption. Traffic noise has been shown to increase the frequency of awakening, increased arousals (i.e. EEG defined awakening) and sleep stage transitions, which cause increased stage 1 sleep, reduced slow wave sleep, and reduced total sleep time (category 1A and 2A) [41,107,120]. Similarly, aircraft flyovers [7] and passing trains [26] have been shown to be associated with increased awakening, increased sleep stage transitions, and reduced slow wave sleep and total sleep time (category 3A). These findings are consistent with the results of an observational EEG study (category 3A) that found that overnight bell tolling from church bells is associated with night waking in a dose response manner [14]. Several survey and quasi-experimental studies using questionnaires, self-report, or actigraphy (category 2B, 2C) confirm that exposure to intermittent noise is associated with sleep disruption. Individuals who are exposed to higher levels of traffic noise self-report more disrupted sleep [10,81,103,114] and increased use of sleeping pills [103] relative to individuals with lower nocturnal traffic noise exposure. Similar findings have been reported for actigraphically-measured sleep disruption from aircraft noise (category 2B) [96] and also questionnaire-based measures of sleep disruption from rail noise (category 3C) [10]. In contrast, a questionnaire-based study (category 3C) comparing modeled aircraft noise on sleep found no changes in sleep insufficiency among those living under higher modeled level noises [34], however, that study used the exposure category of <60 dBA as a comparison group, which would include noises loud enough to be disruptive to sleep. Together, these findings largely confirm that intermittent noise is disruptive to sleep outcomes.

In summary, we reviewed numerous high quality studies that support the notion that noise pollution causes sleep disruption. Our findings suggest that exposure to intermittent noises above 35 dB are associated with reduced sleep quality and quantity. There were few studies describing the influence of continuous noise exposure on sleep outcomes warranting further research in this area. In general, our findings support the importance of locating bedrooms away from common spaces in order to reduce the impact of household and familiar noise on sleep. In buildings where sleep and common spaces must be co-located, such as in hospitals, hotels, and dormitories, measures to reduce noise emanating from other rooms (such as sound attenuating doors) should confer a positive impact on sleep quality and quality for residents. Similarly, bedrooms should be insulated against exposure to noise pollution from the outside environment, particularly when buildings are situated near highways, railways, and airports. In cases where noise pollution cannot be eliminated through insulation or other sound attenuating measures, the use continuous white noise may be useful in minimizing sleep disruption.

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S0360132318300325

Nocturnal road traffic noise: A review on its assessment and consequences on sleep and health

Sandra Pirrera , ... Raymond Cluydts , in Environment International, 2010

SOL is the first variable responsible for a reduction in total sleep time following a noise disturbed night. Exposure to road traffic noise has been related to prolonged SOL (Muzet, 2007; Griefahn and Spreng, 2004, Griefahn and Gros, 1986). Also, Eberhardt et al. (1987) found an increase in SOL with intermittent noise exposure of 55  dB(A). However, many other studies did not find any significant, objectively measured increase in SOL (Eberhardt and Akselsson, 1987; Kawada et al., 1997; Marks and Griefahn, 2007; Kuroiwa et al., 2002; Wilkinson and Campbell, 1984; Vallet et al., 1983; Öhrström and Skånberg, 2004). As contradictory results appear for objective measurements of SOL, the operational definition of SOL is seldom specified. However, it is very relevant as it can range from the time between lights out and the first occurrence of sleep stage N2 or sleep stage N2 prolonged with 5   min or defined according to the Cole–Kripke sleep-scoring algorithm for actigraphy, which is the start of the first 20-min block with more than 19   min of sleep (Kuroiwa et al., 2002; Eberhardt and Akselsson, 1987; Öhrström and Skånberg, 2004). Yet, more stable and consistent findings of the effects of noise exposure on subjective SOL have been observed (see further). More studies using objective sleep measurements are needed to allow firmer conclusions on the effects of noise intrusion on sleep onset latency, a remark which was previously formulated by Griefahn (2002).

Read full article

URL:

https://www.sciencedirect.com/science/article/pii/S0160412010000474

olmedamilefor.blogspot.com

Source: https://www.sciencedirect.com/topics/earth-and-planetary-sciences/intermittent-noise

Share this post