Appendix H. Modelling and Measurement of Underwater Noise (Subacoustech, 2008)

Transcription

1 Appendix H Modelling and Measurement of Underwater Noise (Subacoustech, 2008)

2 Submitted to: Associated British Ports Ocean Gate Atlantic Way Southampton S014 3QN Submitted by: David Lambert Chase Mill Winchester Road Bishops Waltham Hants SO32 1AH Tel: Tel: +44 (0) Fax: Fax: +44 (0) website: website: Modelling and measurement of underwater noise associated with the proposed Port of Southampton capital dredge and redevelopment of berths 201/202 and assessment of the disturbance to salmon. J R Nedwell, S J Parvin, A G Brooker and D R Lambert 05 December 2008 Subacoustech Report No. 805R0444 Approved by Technical Director: Dr J R Nedwell This report is a controlled document. The Report Documentation Page lists the version number, record of changes, referencing information, abstract and other documentation details.

3 Executive Summary This study has been undertaken by., on behalf of Associated British Ports (ABP). It provides an assessment of the underwater noise from construction and dredging operations associated with the proposed capital dredge of the navigational approach channel and the redevelopment of berths 201 / 202 at the Container Terminal. These works are required to provide greater periods of access for deep draughted vessels. Southampton Water, including the 201 / 202 site, form part of the migratory route for fish such as the Atlantic Salmon (Salmo salar) and Sea Trout (Salmo trutta), that move to and from the Rivers Test, Itchen and Hamble. This study uses a combination of underwater noise measurements and underwater acoustic modelling to assess impact zones for piling operations and dredging associated with the proposed developments. For comparison, data are also provided for the current levels of underwater noise in Southampton Water, and for the levels from typical shipping movements. A review of the impact of underwater noise has highlighted the levels of broadband underwater noise that may cause lethality and physical injury in species of fish, as well as the frequency components of underwater noise to which fish are sensitive. Audiogram data, for example, indicates that fish are most responsive to underwater sounds at frequencies from 10 Hz to 1000 Hz. Compared with some other species of fish, the Atlantic salmon and Sea trout are only moderately sensitive to underwater sound. Underwater noise has been monitored throughout Southampton Water and at the berth 201 / 202 site throughout a working day. Broadband (1 Hz to 120 khz) Sound Pressure Levels varied from a minimum of 101 to a maximum of 141 db re. 1 µpa, with mean levels of typically 120 to 130 db re. 1 µpa. These levels of underwater noise are relatively high when compared with published data for deep ocean noise. Perceived levels of underwater noise for the salmon have been assessed by calculating the level of the noise above the published hearing threshold for the species. The level of the background noise in Southampton Water above hearing threshold (db ht ) for the salmon varied from 0 to 38 db ht (Salmo salar). The monitoring at the 201 / 202 berth site indicated an increase in underwater noise related to activities in the harbour and the container terminal, with further increases in underwater noise as both small and large vessels manoeuvred in and around the Upper Swinging Ground. Measurements of underwater noise from backhoe dredging operations have indicated that the highest levels of underwater noise occur when the excavator is in contact with the seabed. Based on a 50 db ht (Salmo salar) criteria (loudness levels below 50 db ht are considered to have a low likelihood of causing a behavioural disturbance) the underwater noise during backhoe dredging operations may cause a behavioural disturbance to salmon to a range of only 15 m. Analysis of the underwater noise from suction dredging operations has indicated that the 50 db ht (Salmo salar) loudness zone may extend to a range of 50 m. Previous measurements have been used to estimate the underwater noise during rock pecker operations above the water line to break up the existing concrete quay wall. The analysis of these results has indicated that the noise during these operations may cause a behavioural disturbance to salmon to ranges of approximately 200 m. Currently, there is no data to provide a more accurate assessment of the noise from this activity, but if the opportunity arises it is intended to obtain more specific measurements. Data from previous impact pile driving operations at this scale has indicated that the estimates of source level noise for pile driving operations proposed for the berth 201 / 202 development exceed levels likely to cause disturbance to fish. Physical injury to fish may occur to a range of approximately 6 m. The perceived levels of noise for the salmon are likely to exceed 50 db ht (Salmo salar) at the quay wall, and in the waters to the shoreline on the opposite side of i

4 Southampton Water, at Marchwood. The underwater noise may therefore cause a disturbance to migratory salmon and sea trout. A review of engineering techniques such as bubble curtains or sound absorbing pile sleeves, often used to reduce the impact of underwater noise, has indicated that these methods are unlikely to provide sufficient low frequency sound attenuation to reduce the perceived noise for salmon to a level that would be acceptable (i.e. Less than 50 db ht (Salmo salar) for half of the river width). ii

5 List of Contents Executive Summary... i List of Contents... iii 1 Introduction Measurement of underwater noise Introduction Units of measurement Measurements Measurement of underwater noise Impact of underwater sound and vibration on fish Introduction Lethality and physical injury US NOAA National Marine Fisheries Service Hastings and Popper (2005) Popper et al, Parvin et al, Behavioural response Frequency weighting db ht (Species) Fish audiogram data Hearing sensitivity of the salmon (Hawkins and Johnston, 1978) Response of biological systems to noise Background noise measurements Introduction Measurements Measurement sites Results of background underwater noise monitoring Southampton Water Measurements of the 7 th May, Measurements of the 4 th April, Measurements of the 2 nd May, Summary of background noise measurements Long term monitoring of underwater noise, berths 201 / Measurement of underwater noise from shipping Introduction Review of published shipping noise data Measurements CMA CGM Verlaine Kyoto Express Vega Stockholm Red Jet Ferry Summary of shipping noise impact on salmon Underwater noise from construction operations during berth 201/202 development Introduction Piling operations Vibro-piling Hydraulic piling Bored (Auger) piling Impact Piling Underwater noise from pile driving operations Characteristics of underwater noise from pile driving operations Source Level noise from impact piling Sound propagation modelling Introduction iii

6 mm diameter tubular steel piling mm tubular steel piling mm width sheet piling Underwater noise from concrete breaking on the existing quay Underwater noise from rock breaking operations below the water line Summary of construction noise impacts on salmon Underwater noise from channel dredging operations Introduction Mechanical dredging Introduction Underwater noise measurements during backhoe dredging operations Results of underwater noise measurements during backhoe dredging Behavioural avoidance Suction dredging Introduction Underwater noise from suction dredging operations Summary of dredging noise impact on salmon Summary and Conclusions References A Calibration data Report Documentation Page iv

7 1 Introduction This study has been undertaken by., on behalf of Associated British Ports (ABP). It provides an assessment of the underwater noise for the proposed construction and dredging operations associated with the capital dredge of the navigational approach channel and the redevelopment of berths 201 / 202 at the Container Terminal. Associated British Ports is undertaking the redevelopment of the existing container facility at berths 201 and 202, which will require the piling of cylindrical steel piles into the estuary bed over a period of at least four months. The work will also involve breaking up the existing concrete quay wall, and dredging of the area in front of berths 201/202 to accommodate deeper draughted vessels. Dredging operations are also proposed to the main Southampton Water navigational channel, both to widen the channel, and to dredge to a depth that will allow greater opportunity for access and passing for deeper draughted vessels. The marine environment within Southampton Water includes the Solent and Southampton Water Special Protection Area (SPA) and Ramsar site, the Solent Special Area of Conservation (SAC), along with component SSSIs. Southampton Water forms part of the migratory route for fish including the Atlantic Salmon (Salmo salar) and Sea Trout (Salmo trutta), that move to and from the Rivers Test, Itchen and Hamble. ABP has submitted a Scoping Study for the proposed construction operations associated with both the capital dredge and the 201 / 202 developments. The Scoping Opinion received from the Marine and Fisheries Agency highlighted the need for a review of background noise levels in Southampton Water, and a detailed assessment of the underwater noise generated during these activities. In particular, the review highlighted the need for an assessment of the likely impact on the migratory passage of the Annex II protected species, the Atlantic salmon (Salmo salar). The review of the Scoping Study by statutory consultees reinforced the requirement for undertaking an assessment of the underwater noise impact of the proposed construction activities. This technical report provides an overview of the metrics for measuring underwater noise and the levels that may cause impacts upon fish. The results of a programme of background underwater noise measurements throughout Southampton Water are presented, together with levels from shipping. Previous measurements of underwater noise from impact piling and construction operations have been used to model, and thereby predict the levels of noise in Southampton Water during the dredging activities and construction operations. Measurements from published literature and underwater noise measurements have been used to quantify the underwater noise expected from suction dredging and mechanical backhoe dredging operations. Critically, all of this noise data is assessed in terms of its likely perception by the Atlantic Salmon, and hence the report identifies those activities that are likely to require further mitigation. Trout have a similar hearing morphology and ability to salmon, and hence the results for salmon may be applied conservatively to trout also. 1

8 2 Measurement of underwater noise 2.1 Introduction This section of` the report introduces some of the metrics used to characterise underwater sound together with some typical examples of their use. Sound travels much faster in water (approximately 1500 m.s -1 ) than in air (340 m.s -1 ). Since water is a relatively incompressible, dense medium, the pressures associated with underwater sounds tend to be much higher than in air. As an example, background levels of sea noise of approximately 130 db re. 1 µpa for UK coastal waters are not uncommon (Nedwell et al, 2003 and 2007a) This level equates to about 100 db re. 20 µpa in the units that would be used in air. Such levels in air would be considered to be hazardous, however, marine animals have evolved to live in this environment and are thus relatively insensitive to sound pressure when compared with terrestrial mammals. 2.2 Units of measurement Sound measurements underwater are usually expressed using the decibel (db) scale, which is a logarithmic measure of sound. A logarithmic scale is used because rather than equal increments of sound having an equal increase in effect, typically a constant ratio is required for this to be the case, that is, each doubling of sound level will cause a roughly equal increase in loudness. Any quantity expressed in this scale is termed a level. If the unit is sound pressure, expressed on the db scale, it will be termed a Sound Pressure Level. The fundamental definition of the db scale is given by Level = 10 x log 10 (Q/Q ref ) eqn. 2-1 where Q is the quantity being expressed on the scale, and Q ref is the reference quantity. The db scale represents a ratio and, for instance, 6 db really means twice as much as.. It is therefore used with a reference unit, which expresses the base from which the ratio is expressed. The reference quantity is conventionally smaller than the smallest value to be expressed on the scale, so that any level quoted is positive. For instance, for sound in air a reference quantity of 20 µpa is usually used, since this is the threshold of human hearing. A refinement is that the scale when used with sound pressure, is applied to the pressure squared, rather than the pressure. If this were not the case, if the acoustic power level of a source rose by say 10 db the Sound Pressure Level would rise by 20 db. So that variations in the units agree, the sound pressure must be specified in units of RMS pressure squared. This is equivalent to expressing the sound as Sound Pressure Level = 20 x log 10 (P RMS /P ref ) eqn. 2-2 For underwater sound, typically a unit of one micropascal (µpa) is used as the reference unit; a Pascal is equal to a pressure of one Newton over one square metre. One micropascal equals one-millionth of this. 2.3 Measurements Sound may be expressed in many different ways depending upon the particular type of noise, and the parameters of the noise that allow it to be evaluated in terms of a biological effect. These are described in more detail below. Peak Level. The peak level is the maximum level of the acoustic pressure, usually a positive pressure. This form of measurement is often used to characterise underwater blast where there is a clear positive peak following the detonation of explosives. Examples of this type of measurement used to define underwater blast waves can be found in Bebb and Wright (1953 and 1955), Richmond et al (1973), Yelverton et al. (1973) and Yelverton (1981). The data from these studies has been widely interpreted in a number of reviews of the impact of high level 2

9 underwater noise causing fatality and injury in human divers, marine mammals and fish (See for example Rawlins (1974), Hill (1978), Goertner (1982), Richardson et al. (1995), Cudahy and Parvin (2001), Hastings and Popper (2005)). The peak sound level of a freely suspended charge of TNT in water can be estimated from Arons (1954), as summarised by Urick (1983). For offshore operations such as well head severance typical charge weights of 40 kg may be used, giving a source peak pressure of 195 MPa or 285 db re. 1 m (Parvin et al. 2007). Peak to peak level. The peak to peak level is usually calculated using the maximum variation of the pressure from positive to negative within the wave. This represents the maximum change in pressure (differential pressure from positive to negative) as a transient pressure wave propagates. Where the wave is symmetrically distributed in positive and negative pressure, the peak to peak level will be twice the peak level, and hence 6 db higher. Peak to peak levels of noise are often used to characterise sound transients from impulsive sources such as percussive impact piling and seismic airgun sources. Measurements during offshore impact piling operations to secure tubular steel piles into the seabed have indicated peak to peak source level noise from 244 to 252 db re. 1 m for piles from 4.0 to 4.7 m diameter (Parvin et al. 2006a, Nedwell et al. 2007a). Sound Pressure Level (SPL). The Sound Pressure Level is normally used to characterise noise and vibration of a continuous nature such as drilling, boring, continuous wave sonar, or background sea and river noise levels. To calculate the SPL, the variation in sound pressure is measured over a specific time period to determine the Root Mean Square (RMS) level of the time varying sound. The SPL can therefore be considered to be a measure of the average unweighted level of the sound over the measurement period. As an example, small sea going vessels typically produce broadband noise at source SPLs from 170 to 180 db re. 1 m (Richardson et al, 1995), whereas a supertanker generates source SPLs of typically 198 db re. 1 m (Hildebrand, 2004). Where an SPL is used to characterise transient pressure waves such as that from seismic airguns, underwater blasting or piling, it is critical that the time period over which the RMS level is calculated is quoted. For instance, in the case of a pile strike lasting say a tenth of a second, the level calculated over that tenth of a second will be 10 db higher (i.e. the apparent energy per second ten times higher) than that taken over one second. Sound Exposure Level. When assessing the noise from transient sources such as blast waves, impact piling or seismic airgun noise, the issue of the time period of the pressure wave (highlighted above) is often addressed by measuring the total acoustic energy (energy flux density) of the wave. This form of analysis was used by Bebb and Wright (1951 to 1955), and later by Rawlins (1987) to explain the apparent discrepancies in the biological effect of short and long range blast waves on human divers. More recently, this form of analysis has been used to develop an interim exposure criterion for assessing the injury range for fish from impact piling operations (Hastings and Popper (2005), Popper et al, (2006)). The Sound Exposure Level (SEL) sums the acoustic energy over a measurement period, and effectively takes account of both the SPL of the sound source and the duration the sound is present in the acoustic environment. Sound Exposure (SE) is defined by the equation: SE T 0 2 p ( t) dt eqn. 2-3 Where p is the acoustic pressure in Pascals, T is the duration of the sound in seconds and t is time. The Sound Exposure is a measure of the acoustic energy and therefore has units of Pascal squared seconds (Pa 2 -s). 3

10 To express the Sound Exposure on a logarithmic scale by means of a decibel, it is compared with a reference acoustic energy level of 1 µpa 2 (P 2 ref) and a reference time (T ref ). The Sound Exposure Level (SEL) is then defined by: T 2 p ( t) dt 0 SEL 10 log eqn P T ref ref By selecting a common reference pressure P ref of 1 µpa for assessments of underwater noise, the SEL and SPL can be compared using the expression: SEL = SPL + 10 log 10 T eqn. 2-5 where the SPL is a measure of the average level of the broadband noise, and the SEL sums the cumulative broadband noise energy. Therefore, for continuous sounds of duration less than 1 second, the SEL will be lower than the SPL. For periods of greater than 1 second the SEL will be numerically greater than the SPL. (i.e. For a sound of 10 seconds duration the SEL will be 10 db higher than the SPL, for a sound of 100 seconds duration the SEL will be 20 db higher than the SPL and so on). In this study, as the background sea noise, ship noise and dredging noise are all continuous sounds, the measurement period T has been normalised to a 1 second period. Hence, for this case, the Sound Exposure Level in db re.1µ Pa 2 -s is numerically equal to the Sound Pressure Level in db re. 1µPa (i.e. 10 log 10 (1) = 0). Particle velocity. The use of particle velocity as an alternative or compliment to sound pressure has been advocated for sound measurements. There is evidence that many species of fish are sensitive to particle velocity rather than pressure (Hawkins, (1981)) Hastings and Popper (2005)). Particle velocity defines the movement of the particles of a medium under the influence of a sound wave. In a free acoustic field, the particle motion is related to the acoustic pressure by the expression; P = v. Z eqn Where v is the particle displacement under the influence of an acoustic pressure P, in a medium with an acoustic impedance Z. It is common to quote the level referenced to the particle velocity of a 1 Pa plane wave. For deep water, this has the advantage that the level of the sound is the same whether quoted in particle velocity or pressure and hence, measurements of acoustic pressure undertaken with a hydrophone can be converted to units of particle velocity.. A note of caution should be applied here, however. A number of recent studies have used this approach to infer particle velocity levels in shallow coastal waters based on pressure measurements undertaken with a hydrophone (See for example Thomsen et al., 2006). For applications in shallow coastal waters this approximation is not appropriate as the pressure measurements are being undertaken in the near field of a pressure release surface (i.e. the water surface). For discussion on this see Kinsler et al., (1982). For these cases the particle velocity must be measured directly. The best approach to this is to use the pressure gradient measurement technique developed by Fahy (1977). It should be noted that particle velocity is a vector quantity, and hence it must be quoted appropriately, say as a magnitude only or in vertical and two horizontal components. Sound propagation. It is conventional, where possible, to evaluate measurements of sound in terms of the effective level of the source (the Source Level, or SL) and the rate at which this 4

11 energy decays with distance (the Transmission Loss, or TL). The use of the SL/TL formulation has the advantage that it decouples the losses during propagation from the strength of the sound source. Source Level. Where there is a single and well-defined source of noise, underwater sound pressure measurements are usually expressed as db re 1 1m. The Source Level is a versatile quantity that can be used, for instance, in estimating the level of sound that that source would generate in a different acoustic environment. However, there is often confusion concerning the concept inherent in Source Level of apparent level at a distance of one metre from the source. In fact, since the measurements are usually made at some distance from the source (in the acoustic far field), and extrapolated back to the source, the true level at one metre may be very different from the Source Level. Indeed, a Source Level may be quoted for sources having dimensions greater than one metre, such that an actual level at one metre cannot be measured. The Source Level may itself be quoted in any of the measures above, for instance, a piling source may be expressed as having a peak to peak Source Level of 200 db re 1 1 metre. Transmission Loss. As underwater sound propagates away from the source it reduces in level. This reduction of sound with range is known as the Transmission Loss, where the Transmission Loss, or TL, is defined as TL P0 20 log eqn. 2-7 P R where P 0 is the acoustic pressure at a point at 1m from the source, and P R is the acoustic pressure at range R away from it. The Transmission Loss is therefore a measure of the rate at which the sound energy decreases. The sound from a source can travel through the water both directly and by means of multiple bounces between the surface and seabed. Sound may also travel sideways through the rocks of the seabed, re-emerging back into the water at a distance. Refraction and absorption further distort the sound. Predicting the level of sound from a source is therefore extremely difficult, and use is generally made of simple models or empirical data based on measurements for its estimation. Propagation modelling. Sound propagation is described by the equation L(r) = SL TL eqn. 2-8 where L(r) is the Sound Pressure Level at distance r from a source (m), SL is the (notional) source level, and TL is the transmission loss (Kinsler et al, 1982). The Transmission Loss is frequently described by the equation TL = N log(r) + r eqn. 2-9 where r is the distance from the source (m), N is a factor for attenuation due to geometric spreading, and is a factor for the absorption of sound in water and boundaries in db.m -1 (Urick, (1983), Kinsler et al, (1982)) By combining expressions 2-8 and 2-9, the level of sound at any point in the waterspace can be estimated from the expression; L(r) = SL N log 10 (r) α r eqn Over short distances absorption effects have little influence on the Transmission Loss and are sometimes ignored. 5

12 2.4 Measurement of underwater noise The fundamental equipment used for measuring underwater noise is relatively simple. Typical measurements will use a hydrophone (underwater microphone) matched to a conditioning amplifier and recording system. However, the measurements can lay great demands on the measurement equipment, caused by the wide frequency range and large dynamic range required to accurately specify underwater noise. In respect of frequency, the hearing range of most fish species is encompassed by a range from about 10 Hz to in excess of 1 khz, (see section 3). Marine mammals hear much higher frequencies, in many cases to well in excess of 100 khz. Although hydrophones are available that cover these frequency ranges, simultaneous measurement of underwater sound over the full audiometric range of fish and marine mammal species places considerable demands on the recording equipment. Frequency range. At a minimum, from the Nyquist Criterion data must be sampled at a minimum of twice the maximum frequency to be recorded. Thus, for marine mammals capable of hearing sound at say 120 khz any waveform must be recorded at a minimum sample rate of 240 samples per second. Much of the information presented in the open literature is recorded to a maximum frequency of ten or twenty khz and is hence completely unsuitable for general analysis. Spectral dynamic range. The issue of dynamic range is often missed. As the noise levels are much higher at the lower frequencies than at the high frequencies, an adequate dynamic range is required, which will enable the high levels of low-frequency noise to be recorded without clipping 1, but also allowing much lower level high frequency sound to be recorded. Typically, the dynamic range of any recording made must be at least 70 db over this range, if the data are to be subsequently analysed to allow the noise impact on most marine animals to be judged. Temporal dynamic range. The situation is worse for measurements of transient pressure waves such as those associated with impact piling, as the dynamic range of the equipment has to be sufficient not only to deal with the spectral dynamic range, but also with changes in level with time caused by the pile being struck etc.. Slew-rate limiting. Another often unrecognised limitation of recording equipment is that of slewrate limiting. Many preamplifiers perform well at low signal levels, but with the high amplitude signals the active devices are unable to supply enough current to drive capacitive loads, such as long cables. Under these circumstances they change from being voltage drivers to constantcurrent drivers, resulting in distortion of the waveform. Unfortunately, much of the public domain data relating to underwater sound and vibration, and particularly that from biological research tends to be of poor quality as a result of these limiting effects. Data must therefore be interpreted with care. All of the underwater noise and vibration measurements undertaken as part of this study where sampled, digitised and stored on a laptop computer system as high frequency digital files (typically 350,000 samples per second). This means that the data can be assessed in any of the noise assessment formats described, or in any other future format required by the regulatory authorities. Subsequent analysis of the acoustic data was conducted over the frequency range from 1 Hz to 120 khz, although for convenience spectral levels of noise in this report are presented over the frequency range from 1 Hz to 100 khz. 1 Clipping occurs when either the instrument or its associated electronic amplification circuitry is overloaded. The signal that is recorded has a characteristic square top, with the measured level being substantially below the actual signal level. This can lead to a gross underestimate of the true magnitude of the signal. 6

13 3 Impact of underwater sound and vibration on fish 3.1 Introduction Over the last 20 years interest in the hearing of marine animals has increased greatly, fuelled by the desire to understand how sound may impact on animals in noisy environments. There have, however, been relatively few studies on determining sound levels that may give rise to impacts on fish. The anatomical, behavioural and physiological variation among fish species is considerable, with the variation in the ear and associated structures of fish species indicating that various species detect and process incident sound and vibration in different ways (Popper and Fay, 1993). Consequently there is a wide variability in sensitivity both in terms of the minimum levels of underwater sound pressure and particle velocity that can be detected, and also the frequency response range for particular species of fish. A number of reviews of the impact of sound and vibration on fish have characterised fish as either hearing specialists or hearing generalists (see, for example, Hawkins (1981), Turnpenny et al. (1994), Popper et al. (2004), Hastings and Popper (2005), Thomsen et al. (2006), Madsen et al. (2006)). Hearing specialists have a high sensitivity to underwater sound and vibration. This is normally attributed to adaptations of the fish morphology that make the species particularly sensitive to sound. The Atlantic herring (Clupea harengus) and sprat (Sprattus sprattus), for example, have an extension of the swim bladder (bulla) that terminates within the inner ear (Blaxter et al., 1981, Popper et al., 2004). These species of fish have both a lateral line sensitive to particle motion and a swimbladder that translates incident pressure into local particle motion. Fish hearing generalists can be divided into those species that contain a swimbladder, and species such as flat fish that do not. For those fish species that posses a swimbladder sensitivity to sound and vibration is related to the proximity of the swimbladder to the inner ear. Cod (Gadus morhua) have an anterior part of the swimbladder that, although not connected to the inner ear, is in close proximity, and hence cod have high sensitivity to sound. In comparison, the Atlantic salmon (Salmo salar), the species of interest in this study, possess a substantial swimbladder, however, as it is not in close proximity to the inner ear, they are less sensitive to underwater sound than some species of fish. Measurements of the hearing of trout have indicated that while they have a similar morphology to salmon they are rather less sensitive (Nedwell 2003). 3.2 Lethality and physical injury Few studies have been carried out in order to determine sound pressure levels that might give rise to fatalities in species of fish. Yelverton et al. (1975) exposed a total of eight species of fish to explosive sound. They were a mixture of fish with a ducted swimbladder (top minnow (Gambusia affinis), goldfish (Carassius auratus), carp (Cyprinus carpio), rainbow trout (Salmo gairdineri), and channel catfish (Ictalurus punctatus)) and those with a non-ducted swimbladder (guppy (Lebiastes reticulates), bluegill (Lepomis macrochirus), and large mouth bass (Micropterus salmoides)). The fish were exposed to blasts having extremely high overpressure and with varying impulse lengths. It was found that fish mortality was correlated to body mass and the magnitude of the impulse of the blast wave. The injuries sustained by the fish included swim bladder rupture, and kidney and liver damage. The blast-mortality model for fish indicates that for a fish of body mass 1 kg an impulse of 340 Pa.s would give rise to 50% mortality. Similarly, no injury would occur to a 10 kg fish exposed to an impulse of 138 Pa.s. Of significance when considering the likelihood of fish mortality from an activity is that the juveniles of a species are more susceptible to impulse type noise than the larger adults of the species. Based on the studies of Yelverton et al (1975), a recent guidance document on susceptibility of fish to the underwater noise from pile driving construction activities (Popper et al, 2006) highlighted the likelihood of mortality from pile driving noise in terms of relative fish size. This also indicates that smaller juvenile fish are more susceptible than larger adults. 7

14 In contrast Engås et al. (1993 and 1996) divided their observations on cod and haddock (Melanogrammus aeglefinus) into two groups according to fish size, above or below 60 cm length. It was discovered that a higher proportion of the larger size-group was dispersed by underwater sound than of the smaller sized fish. The authors could not provide an explanation of this effect, though sensitivity to low sound frequencies may be enhanced in larger fish owing to a lower swimbladder resonant frequency. The work of Yelverton et al. (1975) indicates that, for a given fish mass, there is a peak pressure and impulse threshold below which mortality does not occur but which, nonetheless, might give rise to structural damage of body tissues. However relatively little work has been carried out on fish to determine the incident sound levels at which, for instance, lung or eye damage might arise. It has been observed that fish exposed to pile driving sound have undergone rupture of the swimbladder (Caltrans, 2001). However, rainbow trout and channel catfish, exposed to a received level of 193 db RMS re 1 µpa from the US Low Frequency Active Sonar, showed virtually no damage (Hastings and Popper, 2005). There has been some debate over whether it is the positive or negative part of the overpressure of the sound wave that causes physiological injury. However, Govoni et al. (2003) concluded that the total energy in the sound wave was responsible for the observed effects of underwater explosions on juvenile pinfish. Other non-auditory damage to fish caused by high intensity sound has included capillary rupture in the skin, neurotrauma or brain damage, and eye haemorrhage (Hastings, 1995). Hastings also reported that goldfish did not survive a 2 hour exposure to 250 Hz signals at a received level of 204 db re. 1 µpa (peak), while blue gouramis (Trichogaster trichopterus) did not survive a 30 minute exposure at 150 Hz and a received level of 198 db re 1 µpa (peak). The criteria that have been recommended for assessing fish mortality and physical injury from underwater noise are based on the broadband level of the noise. These are presented below US NOAA National Marine Fisheries Service The US National Oceanic and Atmospheric Administration (NOAA), National Marine Fisheries Service currently quote a peak sound pressure of 180 db re. 1 µpa for determining the extent of physical injury from pile driving operations. Hastings and Popper (2005), however, dismiss this limit, stating that the scientific basis for this value is obscure, if not completely absent. Based on data from Popper et al, 2005, Popper et al 2006 states that the 180 db re. 1 µpa peak level is considerably lower that the sound pressures that could cause injury in fish Hastings and Popper (2005) Hastings and Popper (2005) provided a review of both peer reviewed and unpublished reports on noise exposure of fish, including data on pile driving operations. The study recommends using the Yelverton et al (1975) experimental studies to determine lethal (50%) and physical injury (0% lethality) injury models, and also discuss auditory injury criteria. The auditory injury criteria are based on tissue damage in fish hearing specialists (i.e. Fish with a swimbladder) following exposure to underwater sound at 120 db above the fish hearing threshold following a 1 hour exposure, and hearing impairment following a 1 hour exposure to sound at 100 db above hearing threshold. (In terms of noise exposure to the salmon, the audiogram data of Hawkins and Johnson (1978) (See Figure 3.3) indicates that the species is most sensitive to underwater sound at a frequency of 160 Hz, where the threshold Sound Pressure Level is 95 db re. 1 µpa. Based on this data, underwater noise might cause tissue damage to the auditory system (PTS) of the salmon following 1 hour exposure at a level of 215 db re. 1 µpa. Hearing impairment (TTS) might occur following exposure at a level of 195 db re. 1 µpa, for a period of 1 hour). For assessing the impact of pile driving operations the Hastings and Popper (2005) study recommends using the Sound Exposure Level (SEL), rather than the peak pressure of the noise. As discussed in section 2.3 the SEL is the sum of the acoustic energy and therefore has units of 8

15 Pascal squared seconds (Pa 2 -s). Where the sound is measured over a one second averaging period (see section 2.3), the one second SEL in db re: 1 μpa 2 sec can be shown to be numerically equal to the (RMS) Sound Pressure Level of the noise in units of db re.1 µpa Popper et al, 2006 Popper et al, (2006) present an interim guidance for protecting fish from injury from the underwater noise from pile driving operations. Rather than replace the use of peak sound level as the parameter for indicating the extent of the injury zone during pile driving operations (as proposed by Hastings and Popper (2005), the interim guidance proposes the use of both the peak sound level and the SEL of the noise. The guidance is based on interpreting the data of Yelverton et al (1975), for mortality and injury of fish exposed to blast, but extending the SEL criteria to fish weighing 0.01 g. The guidance is somewhat confused in that it quotes a received level exposure; an SEL of 187 db re.1µ Pa 2 -s and a peak sound pressure of 208 db re.1 µpa as measured 10 m from the source. An assumption has been made here that the levels quoted are received levels (and not those measured or estimated at 10 m from the source). Were measurements to be made at a range of 10 m, as recommended by the guidance, the measuring equipment would be in the very near field of the source (pile, ship or other), and considerable temporal and spatial variation would be expected. Measurements of this type should be made in the far field, but then also quoted as a source level estimate (level at 1 metre) by using a suitable underwater sound propagation model. It is therefore assumed that the Popper et al, 2007 interim guidance proposes a criteria for single strike pile-driving of a maximum received SEL of 187 db re: 1 μpa 2 -sec and a peak received sound pressure of 208 db re 1 μpa peak (equivalent to a peak to peak level of 214 db re.1 µpa) Parvin et al, 2007 As part of study for the UK government department of Business Enterprise and Regulatory Reform (BERR), Parvin et al, (2007) published a review of the impact of high level underwater sound on marine mammals, fish and human divers. The impact criteria proposed was based on studies by Yelverton et al, (1973, 1975 and 1981), historic studies by the UK MOD (see for example Bebb and Wright (1954), Rawlins (1974)), Young (1991), Cudahy and Parvin (2001) and Hastings and Popper (2005). The report discusses injury and fatality from underwater transient pressure waves related to both the peak pressure, and the duration that the peak pressure acts upon the body (the impulse). In terms of a peak pressure level exposure the study indicated that; At incident peak underwater sound levels of 10 MPa ( 260 db re. 1µPa), and above always lethal. At incident peak underwater sound levels of 1 MPa ( 240 db re. 1µPa) increasing likelihood of death or severe injury leading to death in a short time. At incident peak underwater sound levels of 0.1 MPa ( 220 db re. 1µPa) Direct physical injury to gas-containing structures and auditory organs may occur, particularly from repeat exposures. This criteria was presented and used to assess lethal and physical injury zones in a peer reviewed report on the impact piling operations proposed for the construction of offshore wind farms in the Thames Estuary (Shepherd et al, (2006), Parvin et al, (2006a)). It has also been used as a criteria for assessing impact from pile driving operations during UK offshore wind farm developments (Nedwell et al, (2007), Parvin et al, (2006b)). 3.3 Behavioural response A number of studies have noted that changes in fish behaviour may arise following exposure to relatively low level sounds. Engås and Lokkeborg (2002) observed a reduction in the catch of 9

16 haddock and cod that lasted for several days after they had been exposed to seismic airgun emissions. It was suggested that the sound probably caused the fish to leave the insonified area, although there was no data to support this. Slotte et al. (2004) found broadly similar results for blue whiting and herring. Skalski et al. (1992) found that the catch of rockfish reduced by 52% following exposure to a single emission of an airgun at db re 1 µpa (mean peak level). They also observed a startle reaction in fish at levels of sound as low as 160 db re 1 µpa, although catch rates were not subsequently affected. Wardle et al. (2001) noted a similar startle reaction in fish. Cod, pollack and saithe were exposed to airgun emissions of varying intensities at levels above 195 db re 1 µpa and, in each case, the fish made an involuntary reaction in the form of a C start reflex. A comprehensive study of the effects of underwater sound on fish and other marine animals was undertaken by Turnpenny et al. (1994). The study considered levels of sound from seismic sources and pure tone sources that might give rise to fatalities, less serious injuries, deafness and changes in behaviour. Reaction thresholds for the trout and whiting were found to be around 170 db re 1 µpa, although in the bass there was a significant avoidance to the sound at levels above 130 db re 1 µpa. Data on fish avoidance to underwater sound is also available from the use of fish deflection systems, developed to reduce fish kill at power station water inlets (Maes et al., 2004). In this case underwater sound was deliberately introduced to move fish away from the water inlets, preventing them from being killed. The ability of the sound system to remove fish from the area was found to be species dependent, and related to the underwater hearing sensitivity of the species, i.e. comparatively fewer fish that have sensitive hearing were washed into the power station inlet when the acoustic deflection system was operated. For example, for fish species that are comparatively sensitive to underwater sound such as the herring (Clupea harengus) and the sprat (Sprattus sprattus) average intake rates decreased by 94.7% and 87.9% respectively (Maes et al (2004)), indicating that fish were avoiding the high sound field surrounding the power station water inlet. The same data indicated that for the fish species that were considered less sensitive to underwater sound (based on hearing threshold data) only a moderate response to the sound was demonstrated. The efficiency for the flatfish species, the flounder (Platichthys flesus) was at 37%, and for the sole (Solea solea) was at 47%. This data indicates that when considering the behavioural avoidance response of fish species to activities such as construction, it is the perceived level of loudness and vibration of the sound by the species that is important. Consequently, many researchers are now advocating the use of frequency weighting scales to determine the level of the sound in comparison with the auditory response of the aquatic or marine animal. Madsen et al. (2006), for example, recommend that as the impact of sounds impinging on the auditory system is frequency-dependent, noise levels should (as for humans) ideally be weighted with the frequency response of the auditory system of the animal in question. A recent study published by the UK government department of Business Enterprise and Regulatory reform (Nedwell et al, 2007) assessed the data from Maes et al, 2004, and related other published data on the impact of underwater sound upon marine species to the level above hearing threshold or db ht (Species). This frequency weighting technique compares the measured underwater noise with the hearing threshold of a fish or marine mammal species, and provides a db value that represents the level of the sound above the hearing threshold of the species. The metric is therefore analogous with the db(a) scale used to assess the impact of noise for humans in air. The db ht (Species) technique is described in more detail in section 3.4 below. Based on a large body of measurements of fish avoidance of noise (Maes et al., 2004), the following assessment criteria have been published by the UK Government Department of BERR for assessing the potential impact of the construction noise and vibration (Nedwell et al., 2007b): 90 db ht (Species) strong avoidance reaction by most individuals. 10

17 75 db ht (Species) - mild avoidance reaction occurs in a majority of individuals db ht (Species) low likelihood of disturbance. In human perception terms 0 dbht,(i.e. 0 db(a)) represents the threshold of hearing, and 30 dbht for humans (i.e. 30 db(a)) would be typical of the level of noise in a quiet library. A noisy office might have a level of 60 db(a)) 3.4 Frequency weighting db ht (Species) As indicated in the preceding section, the same underwater sound will affect marine species in a different manner depending upon the hearing sensitivity of that species. The measurements of noise in this study have therefore also been presented in the form of a db ht level for the species. The db ht metric or scale incorporates the concept of loudness for a species. The metric incorporates hearing ability by referencing the sound to the species hearing threshold, and hence evaluates the level of sound a species can perceive, rather than its absolute level. This is illustrated in Figure 3-1, which illustrates the same noise spectrum as it is perceived by fish or marine mammals having different hearing ability. The diagram illustrates conceptually the hearing thresholds of three different species, overlaid on a spectrum of noise. The component of the noise that can be heard by each species is represented by the hatched region, which is different for each species. In the case shown, Fish 1 has the poorest hearing (highest threshold) and only hears the noise over a limited low frequency range and at a low level above its threshold. Fish 2 has very much better hearing and hears the main dominant components of the noise. Although having the lowest auditory threshold (that is, the greatest sensitivity to sound), the marine mammal only hears the very high components of the noise and so it may be perceived by this species as relatively quiet. Figure 3-1. Illustration of perceived sound level (dbht) for fish and marine mammal species. Since any given sound will be perceived differently by different species having differing hearing abilities, the species name must be appended when specifying a level. For instance, the same sound might have a level of 70 db ht (Gaddus morhua) for a cod and 40 db ht (Salmo salar) for a salmon. The perceived noise levels of sources measured in db ht (species) are usually much lower than the unweighted (linear) levels, not only because the sound will contain frequency components that the species cannot detect, but also because most aquatic and marine species have high thresholds of perception to (are relatively insensitive to) sound. 11

18 3.5 Fish audiogram data Figure 3-2 presents a comparison of hearing threshold levels for various species of sea and riverine fish. The data has been compiled from a number of peer reviewed publications. In general, fish hear underwater sound over a low frequency range from approximately 10 Hz to 1000 Hz, although the data indicates a considerable variability in sensitivity to underwater sound depending upon the species. The anatomical, behavioural and physiological variation among fish species is considerable, with the variation in the ear and associated structures of fish species indicating that various species detect and process incident sound and vibration in different ways (Popper and Fay (1993)). Consequently, there is a wide variability in sensitivity both in terms of the minimum levels of underwater sound pressure and particle velocity that can be detected, and also the frequency response range for particular species of fish. A number of reviews of the impact of sound and vibration on fish have characterised fish as either hearing specialists or hearing generalists (see for example Hawkins (1981), Turnpenny et al (1994), Popper et al, (2004), Hastings and Popper (2005), Thomsen et al, (2006), Madsen et al (2006)). Hearing specialists have a high sensitivity to underwater sound and vibration. This is normally attributed to adaptations of the fish morphology that make the species particularly sensitive to sound. The herring (Clupea harengus) and sprat (Sprattus sprattus), for example, have an extension of the swim bladder (bulla) that terminates within the inner ear (Blaxter et al (1981), (Popper et al, 2004)). The audiogram data for the herring from Enger (1967) illustrated in Figure 3-2 highlights that the herring is able to perceive underwater sound over a wide frequency range from 30 Hz to 4 khz. Threshold Sound Pressure Level for this species is at approximately 75 db re 1μPa over the range of frequencies from 30 Hz to 1 khz. Fish hearing generalists can be divided into those species that contain a swimbladder, and species such as flat fish that do not. For those fish species that posses a swimbladder sensitivity to sound and vibration is related to the proximity of the swimbladder to the inner ear. Cod (Gadus morhua) have an anterior part of the swimbladder that, although not connected to the inner ear, is in close proximity, and hence cod have high sensitivity to sound. The audiogram data for the cod of Chapman and Hawkins (1973), indicates that it is sensitive to sound over the frequency range from 50 Hz to 400 Hz, where sensitivity varies from 75 to approximately 80 db re 1μPa. Figure 3-1 indicates that other deep water shoaling fish such as the pollack and haddock have a similar auditory range and sensitivity to that of the cod. In comparison, the Atlantic salmon (Salmo salar) possess a substantial swimbladder but, as it is not in close proximity to the inner ear, they are therefore less sensitive to underwater noise and vibration. The audiogram data for the salmon of Hawkins and Johnson (1978) illustrated in Figure 3-1 indicates that the salmon has a peak sensitivity to underwater sound at a frequency of 160 Hz, at a level of 95 db re 1μPa, which coincides with much of the main energy band of piling noise. 3.6 Hearing sensitivity of the salmon (Hawkins and Johnston, 1978) The hearing of the salmon (Salmo salar) was studied by Hawkins and Johnston (1978) by means of a cardiac conditioning technique. Fish were trained to show a slowing of the heart, on hearing a sound, in anticipation of a mild electric shock applied later. The minimum sound level to which the fish would respond was determined for a range of pure tones, both in the sea, and in the laboratory. The preparation of the fish involved a trained response to underwater sound signals. After the subject had been trained, tones were presented, beginning at a high level and then reducing in 3 db steps after each positive response, until the fish no longer responded. At this point the level was increased until the subject again responded, and then reduced again. This reduction and increase of the sound level was carried out until a plateau was reached, and the threshold level was estimated such that the fish could be expected to respond to approximately 50% of presentations. 12

19 The sound to which the fish were exposed was generated by a sound projector located at the same depth as the subject. A total of 4 projectors were used located at distances of 0.65 m, 1.25 m, 2.2 m and 2.7 m from the front of the cage containing the subject. The projectors were held at the required depth by sub-surface floats. The experiments were conducted with five Atlantic salmon, ranging in length from 32 cm to 36 cm, although much of the analysis in the paper considers the results from only four fish. The results of the study are shown in Figure 3-3, as mean Sound Pressure Levels at each test frequency. This is the format that in which the results are usually quoted in terms of threshold of hearing for the species. The paper also calculates the particle displacement using the expression given by Harris (1964). Table 3-1 provides a summary of the variance (Square of the standard deviation) for the datasets of measured sound pressure and calculated particle displacement at each test frequency. The variance in the data increases considerably toward the lower and higher end of the audio range for the salmon. Frequency (Hz) Particle displacement Sound Pressure Table 3-1. Comparison of the variances of sound pressure and particle displacement at threshold for the salmon (data reproduced from the study of Hawkins and Johnstone, 1978) Figure 3-4 presents the variation in measured sound pressure data at increasing range from the sound projector. From this it is clear that there was a wide spread of threshold values for the low frequency tones, the spread depending on the distance between the projector and the subject, the highest thresholds being obtained at the larger separations. It should be noted, however, that all of the measurements were undertaken in the very near field of the sound projectors and hence in a region where very high particle displacements would be expected. The paper concludes that the salmon responded only to low frequency tones (below 380 Hz), and particle motion, rather than sound pressure, as the relevant stimulus. The sensitivity of the fish to sound was not affected by the level of sea noise under natural conditions. Sound measurements made in the River Dee, near Aberdeen, lead to the conclusion that salmon are unlikely to detect sounds originating in air, but that they are sensitive to substrate borne sounds. Compared with the carp and cod the hearing of the salmon was considered to be poor, and more similar to that of fish such as perch and plaice. 3.7 Response of biological systems to noise In considering the results of this study it should be borne in mind that the engineering acoustic units used to quantify a biological response to sound provide a probabilistic estimate of the range of effect. Typically, a range will be provided in metres from the source at which a strong avoidance reaction will occur. This may be defined as the range at which there is avoidance by say 90% of individuals, or equivalently as a 90% likelihood of avoidance by a given individual. This does not mean that no reaction will occur at greater ranges, but rather that the probability of an avoidance reaction will fall below this level. Just as for airborne noise assessment for humans, some people may find a sound unpleasant at a given level and move away from its source, whereas others are unaffected by the same level of noise. The circumstances of the exposure are also important. For someone already in a high noise environment, such as a noisy workshop, a small amount of additional noise may be of no concern. The same person in a very quiet environment such as a library might find the same noise extremely bothersome. At a 13

20 population level, the fundamental unit that the airborne acoustic engineer would apply in all cases for human reaction to noise is the db(a), which analogously to the db ht approximates to a measure of the level of the sound above the hearing threshold for a human in air. When assessing the impact of underwater noise on marine species, either in terms unweighted levels to assess physical injury, or by applying a frequency weighting scale to provide a measure of the underwater sound above hearing threshold (the db ht (Species)), the same caveat applies. Whereas one individual of a species may respond, another may not. Factors such as sex, age, size, feeding and reproductive state, etc., can influence and modify the likely behaviour of any individual of a species to underwater sound at a particular loudness level. As with the db(a), the analysis is intended to indicate the estimated percentage of the species exhibiting a given behaviour at a population level, or the probability of a given response of an individual of the species. 14

21 Figure 3-2. Comparison of hearing threshold for various species of fish Figure 3-3. Hearing threshold levels for the Atlantic salmon (Salmo salar) published by Hawkins and Johnstone (1978). 15

22 Figure 3-4. Variability of hearing threshold data for the Atlantic salmon (Salmo salar) published by Hawkins and Johnstone (1978). 16

23 4 Background noise measurements 4.1 Introduction It follows from the preceding discussions regarding the behavioural effects of sound that a measure of the pre-existing ambient noise in an environment is of critical importance when assessing the impact of noise and vibration from a new activity. In the context of this study, a detailed series of measurements of background levels of underwater noise in Southampton Water has been undertaken, and is provided as a baseline for assessing the impacts of underwater noise from the activities associated with the capital dredge and the berth 201 / 202 development. Background noise can be highly variable and usually comprises a broad range of individual sound sources, some of which are natural and some of which are man-made. The sources of noise may be distant, such as that from shipping, or may be close to the receiver, such as that from waves breaking. These sounds combine to produce an overall continuum of pre-existing noise that defines the noise environment in which marine animals live. It is therefore of critical importance to any assessment of the impact of noise and vibration from an activity to measure and assess the background noise environment. In fast flowing rivers, coastal environments, and busy estuaries the levels of underwater noise will be high compared with typical airborne measurements, or indeed other deep sea underwater noise measurements. If the noise from an activity is below the pre-existing background levels for the region then the sound will be masked, and cannot have an impact on fish or other marine species. Figure 4-1 summarises ambient sea noise data from a deep ocean environment as compiled by Wenz (1962), based on measurements in the eastern Pacific. Similar levels and sources of underwater noise are described by Knudsen (1948). Low frequency ambient noise from 1 to 10 Hz is mainly comprised of turbulence and hydrostatic pressure fluctuations from surface waves and the motion of water at boundaries. It exhibits a dependence on both wind strength and water currents. Between 10 and 100 Hz distant anthropogenic noise dominates, with its greatest contribution between 20 Hz and 80 Hz. The noise in this region is not generally attributable to one specific source, but a collection of sources at distance from the receiver. Distant shipping traffic is the greatest contributor to man-made ambient noise in this frequency band, with received levels from typically 50 to 90 db re. 1 µpa 2.Hz -1, where shipping is heavy. At frequencies above 100 Hz, the ambient noise level depends on weather conditions, with wind and wave related effects creating sound. Sea noise in this high frequency range has been shown to be related to the wind speed, measured using the Beaufort Scale (Knudsen, (1948)), with spectral levels ranging from 20 db re 1 µpa 2.Hz -1 to 80 db re 1 µpa 2.Hz -1. The level of windrelated noise decreases with increasing frequency above approximately 500 Hz, falling with a slope of between 5 and 6 db per octave (doubling of frequency). At frequencies above 20 khz, measured sound levels may be influenced by thermal noise at the lowest of ambient noise levels. The data from Wenz (1962) and Knudsen (1948) are generally accepted as providing an indication of the range of sea noise levels and the source of the dominant noise in each frequency range. However, when considering the spectral levels of ambient sea noise presented by Wenz and Knudsen, it is important to note that these measurements were undertaken over 40 years ago and in very deep water environments compared to the shallow coastal waters and estuarine environments around the UK. The recent review of underwater noise by Hildebrand (2004) cites the data of Mazzuca (2001) in deriving an overall increase of 16 db in low frequency noise during the period from 1950 to the year 2000, corresponding to a doubling of noise power (3 db increase) in every decade for the past five decades. 17

24 Spectrum Level (db re 1 Pa 2.Hz -1 ) 4.2 Measurements. Figure 4-1. Background sea noise levels [adapted from Wenz (1962)]. This section of the report presents the results of an underwater noise measurement programme in Southampton Water, together with estimates of the perceived level of background underwater noise for the salmon. To provide a direct comparison of the underwater noise from the proposed dredging and construction operations during the capital dredge and berth 201 / 202 development, background underwater noise measurements were undertaken throughout the Southampton Water navigation channel. Underwater sound recordings were made on the 4 th April, the 2 nd May, and the 7 th May 2008, at five measurement sites. The study therefore provides an indication of both the temporal and spatial variation in underwater noise in the Southampton Water navigation channel. All of the underwater noise measurements reported here are based upon recordings taken using a Bruel & Kjaer Type 8106 hydrophone, connected to a proprietary Subacoustech hydrophone power supply / amplifier. This provided power to the hydrophone and signal conditioning and 18

25 amplification of the acoustic signal. Measurements presented in this study are based on analysis over the frequency range from 1 Hz to 120 khz. A full description of the analysis equipment and procedures is provided in section 2.4 and Appendix A. 4.3 Measurement sites Underwater noise recordings were undertaken at five sites throughout Southampton Water, from the Container Terminal site at the western end of the Southampton Water channel, to Calshot Bay at the entrance to Southampton Water. The measurement locations are described in more detail below. 1. Bury Swinging Ground, position ( 'N 'W), in the deep water turning ground opposite berths 205 / 206, Southampton Container Terminal. 2. Upper Swinging Ground, position ( 'N 'W), in the deep water turning ground opposite berths 201 / 202, Southampton Container Terminal. 3. Dock Head, position ( 'N 'W), at the end of the Eastern Docks, Grain Terminal, in front of berth Netley Abbey, position ( 'N 'W), at the southern edge of the main shipping channel opposite Netley Abbey. 5. Hook Buoy, position ( 'N 'W), Calshot Bay at the edge of the main shipping channel at the entrance to Southampton Water. 4.4 Results of background underwater noise monitoring Southampton Water Measurements of the 7 th May, The measurements of background underwater noise on the 7 th May 2008 were undertaken during a period from 12:29 to 16:33 hours. The weather conditions were bright and sunny, and were breezy with a slight chop on the water. High water at Southampton was at 12:43, with low water at 18:43. The initial measurements at the Hook Buoy, Calshot Bay were therefore undertaken at high water, with the subsequent measurements undertaken throughout the ebb tide. Figure 4-2 presents a 30 second time history of the underwater noise measured at the Dock Head at 14:13:56. The time history indicates that the noise is rather featureless, reaching peak to peak levels of approximately 60 Pa (156 db re. 1 µpa). The corresponding one second, RMS sound pressures during this period varied from 8.5 to 10 Pa, giving RMS Sound Pressure Levels from to 140 db re. 1 µpa. Figure 4-3 presents a 30 second time history of underwater noise measured at 13:18:19 hours on the 7 th May 2008, at the Netley Abbey measurement position. The figure is presented on the same axes as the previous noise time history (Figure 4-2) to provide a visual indication of the variation in measured underwater noise. For this case, the peak to peak levels of noise are at approximately 10 Pa (140 db re. 1 µpa). The corresponding one second, RMS sound pressures during this period varied from 1 to 1.4 Pa, giving RMS Sound Pressure Levels from 120 to 123 db re. 1 µpa. Figure 4-4 presents the distribution of Sound Pressure Level measured at each of the sites on the 7 th May. The data presented for Hook Buoy, for example, is a summary of the analysis of 280 (n = 280) one second data files at the site. The data for this measurement position indicating that the Sound Pressure Level varied from a minimum of 113 to a maximum of 128 db re. 1 µpa. The distribution of Sound Pressure Level data highlights that the underwater noise across all five measurement sites varied from a minimum of 113 db re. 1 µpa at the Hook Buoy, Calshot Bay, to a maximum of 141 db re. 1 µpa at the Dock Head. The measurements indicate that in terms of the broadband Sound Pressure Level, the Dock Head was the noisiest site. The mean level of 19

26 noise at each site varied from 118 db re. 1 µpa at Netley Abbey, to 134 db re. 1 µpa at Dock Head. The variation in the measured data across each of the five sites, on the three days of underwater noise recording is also summarised in Table 4-1. This presents the underwater noise data both in terms of the variation in Sound Pressure Level and the RMS db ht level for the salmon. Figure 4-5 provides a comparison of the spectral levels of underwater noise in each of the measurement positions. The spectral levels of underwater noise below 10 Hz are typical of those in many coastal regions. The main contribution to the underwater noise is from frequencies in the range from 10 Hz to 1000 Hz, where the spectral levels vary from 80 to 120 db re. 1 µpa 2.Hz -1. There is also some evidence of tonal noise at low frequencies from 10 Hz to 100 Hz. These levels of underwater noise are high compared with those quoted by Wenz (1962) and Knudsen (1948) for levels of underwater noise in the ocean, and therefore indicate the ambient underwater noise in Southampton Water is relatively high. In common with the broadband Sound Pressure Level data summarised in Figure 4-4, the data indicates that the spectral levels of noise were highest at the Dock Head. Noise at frequencies above 1000 Hz also contributes to the overall level, but with spectral levels decreasing at over 25 db per decade to a frequency of approximately 80 khz. The tonal in the Dock Head data centred at a frequency of approximately 55 khz may be due to the noise from a vessel echo sounder. The most important data for supporting the assessment of underwater noise on migratory salmon is the frequency weighted data in terms of the salmon hearing perception. Figure 4-6 presents the summary of the analysis of the background underwater noise data in terms of the perceived loudness by the salmon. The db ht (Salmo salar) data provides a measure of the level of broadband noise above the hearing threshold of the salmon. The distribution of results presented in Figure 4-6 highlights that on the 7 th May the ambient sea noise throughout Southampton Water was perceived at levels from 8 to 38 db ht,(salmo salar), with mean values that varied from 15 db ht,(salmo salar) at Hook Buoy, Netley Abbey and the Upper Swinging Ground, to 34 db ht,(salmo salar) at Dock Head. The measurements undertaken in the Upper Swinging Ground, opposite berths 201 / 202 indicate levels for the salmon that varied from 10 to 20 db ht,(salmo salar) Measurements of the 4 th April, The measurements of background underwater noise on the 4 th April 2008, were undertaken during a period from 06:45 to 11:00 hours. The weather conditions on the 4 th April were initially foggy and still with little or no wind. Later in the morning the fog cleared and it became bright and slightly overcast. Generally, the weather conditions were very calm with minimal wind speeds. High Water on the 4 th April 2008 was at 10:19 and Low Water at 16:19 hours. The initial background noise measurements were undertaken in the Bury and Upper Swinging Ground, with measurements later in the morning toward the eastern end of Southampton Water and the Hook buoy at Calshot Bay. Figures 4-7, 4-8 and 4-9 present the distribution of Sound Pressure Levels, spectral levels of underwater noise and distribution of db ht,(salmo salar) levels respectively for the underwater noise recorded at each of the five measurement positions. In contrast to the data captured on the 7 th May, the data for the 4 th April indicates the highest levels of underwater noise at the Hook Buoy, Calshot Bay. The distribution of measured noise in Figure 4-7 indicates that the broadband Sound Pressure Levels where consistently high, varying from 135 to 139 db re. 1 µpa, with the perceived sound levels for the salmon from 32 to 37 db ht,(salmo salar). The spectral levels of noise illustrated in Figure 4-8 indicate that the increase in noise in this region, compared with the other sites, occurred at frequencies from 20 Hz to 400 Hz. The measurements undertaken in the Upper Swinging Ground, opposite berths 201 / 202 indicate levels for the salmon that varied from 17 to 20 db ht,(salmo salar) on the 4 th April

27 4.4.3 Measurements of the 2 nd May, The measurements of background underwater noise on the 2 nd May 2008 were undertaken during a period from 12:45 to 16:15 hours. The weather conditions on the 2 nd May were bright and sunny, with a breeze of 2 to 3.5 m.s -1, and the water was choppy. High Water on the 2 nd May was at 08:52 and again at 21:22 hours, with low water at 14:51. The initial background noise measurements were undertaken at Dock Head, with measurements later in the afternoon in the Upper and Bury Swinging Grounds, before moving to the Hook buoy and Netley Abbey Figures 4-10, 4-11 and 4-12 present the distribution of Sound Pressure Levels, spectral levels of underwater noise and distribution of db ht,(salmo salar) levels respectively for the underwater noise recorded at each of the five measurement positions on the 2 nd May. On this day, there were three large container vessels on berths 204 to 207 at the Southampton Container Terminal, all of which were undergoing loading / unloading. Consequently, the underwater noise data indicates the highest levels of noise for the measurement position in the Bury Swinging Ground. The underwater noise recorded (n = 115) during this period was extremely consistent with broadband Sound Pressure Levels from 131 to 132 db re. 1 µpa, and levels for the salmon in this region from 25 to 28 db re. 1 µpa.the spectral analysis in Figure 4-11 indicates that the increase in noise in this region, compared with the other sites, occurred at frequencies from 80 Hz to 2000 Hz. In contrast, the levels at the Netley Abbey site on this day varied from 0 to 12 db ht,(salmo salar). The measurements undertaken in the Upper Swinging Ground, opposite berths 201 / 202 indicate levels for the salmon that varied from 11 to 18 db ht,(salmo salar) on the 2 nd May Summary of background noise measurements Measurements throughout Southampton Water have indicated that the underwater noise environment is high compared with levels typically quoted for open ocean noise. Broadband (1 Hz to 120 khz) Sound Pressure Levels varied from a minimum of 101 to a maximum of 141 db re. 1 µpa, with mean levels of typically 120 to 130 db re. 1 µpa. The perceived levels for the salmon varied from 0 to 38 db ht (Salmo salar) which is considerably below the levels likely to have a disturbance effect on the species. There was no consistent trend in the data indicating both the temporal and spatial variability of underwater noise in a busy harbour and estuary environment and it is concluded that current background noise levels do not adversely impact on salmon. 21

28 Sound Pressure Level (db re. 1 µpa) RMS db ht (Salmo salar) 4 th April 2008 Max Min Mean Max Min Mean Bury Swinging Ground [n=95] Upper Swinging Ground [n=155] Dock Head [n= 65] Netley Abbey [n=95] Hook Buoy [n=90] nd May 2008 Bury Swinging Ground [n = 115] Upper Swinging Ground [n = 140] Dock Head [n = 303] Netley Abbey [n = 300] Hook Buoy [n = 175] th May 2008 Bury Swinging Ground [n = 179] Upper Swinging Ground [n = 235] Dock Head [n = 140] Netley Abbey [n = 150] Hook Buoy [n = 280] Table 4-1. Summary of measurements of background underwater noise at five sites in Southampton Water (n = number of measurements) 22

29 Figure 4-2. A 30 second time history of underwater noise recorded at the Dock Head measurement position on 7 th May [File: ]. Figure 4-3. A 30 second time history of underwater noise recorded at the Netley Abbey measurement position on 7 th May [File: ] 23

30 Figure 4-4. Summary of one second RMS Sound Pressure Levels of underwater noise recorded at the 5 measurements positions in Southampton Water, 7 th May The data are presented as a distribution of measured noise levels. Figure 4-5. Comparison of spectral levels of underwater noise at five measurement positions in Southampton Water, 7 th May

31 Figure 4-6. Summary of one second, RMS db ht (Salmo salar) levels at each of the five measurement positions in Southampton Water, 7 th May The data are presented as a distribution of measured noise levels Figure 4-7. Summary of one second, RMS Sound Pressure Levels of underwater noise recorded at the five measurement positions in Southampton Water, 4 th April The data are presented as a distribution of measured noise levels 25

32 Figure 4-8. Comparison of spectral levels of underwater noise at five measurement positions in Southampton Water, 4 th April Figure 4-9. Summary of one second, RMS db ht (Salmo salar) levels at each of the five measurement positions in Southampton Water, 4 th April The data are presented as a distribution of measured noise levels. 26

33 Figure Summary of one second, RMS Sound Pressure Levels of underwater noise recorded at the five measurement positions in Southampton Water, 2 nd May The data are presented as a distribution of measured noise levels. Figure Comparison of spectral levels of underwater noise at five measurement positions in Southampton Water, 2 nd May

34 Figure Summary of one second,rms db ht (Salmo salar) levels at each of the five measurement positions in Southampton Water, 2 nd May The data are presented as a distribution of measured noise levels. 28

35 4.6 Long term monitoring of underwater noise, berths 201 / 202. To understand and quantify the underwater noise environment in the region of berths 201 / 202, the underwater noise was monitored throughout a typical operational working day at the container port. The underwater sound was recorded using identical equipment to that used for the background underwater noise measurements described earlier. For this form of monitoring, however, each one second of underwater noise was processed in real time, with only the broadband Sound Pressure Level and RMS db ht (Salmo salar) levels being stored for subsequent analysis. The long term monitoring at the berth was undertaken on the 10 th June 2008, commencing at 06:15. For these measurements the hydrophone was deployed a distance of 5 m from the quay wall, at the mid-point of the 201 / 202 berths. The results of the monitoring are presented in Figure 4-11, with supporting information on activity at the container port provided in Table 4-2. Initially, at 06:15, there was little or no activity in the harbour. Figure 4-11 indicates that the Sound Pressure Levels during this period varied from 107 to 109 db re. 1 µpa, with perceived levels for the salmon below 0 db ht (Salmo salar). At 06:26 the Queen Victoria, a cruise ship of 90,000 tons, moved up through Southampton Water and manoeuvred into berth 106, at a range of approximately 1000 m from the monitoring position. During this period the broadband Sound Pressure Level at the monitoring position increased from 110 to a maximum of 137 db re. 1 µpa. The perceived levels for the salmon varied from 5 to 35 db ht (Salmo salar). Other notable increases in underwater noise occurred during the period from 09:20 to 09:40 when a support barge was operating alongside the vessel Queen Victoria at berth 106 (range of approximately 1000 m from the measurement position). During this period the broadband Sound Pressure Level was recorded at 127 db re. 1 µpa, with the perceived levels for the salmon at 19 to 21 db ht (Salmo salar). From 10:22 to 10:35 the Margareta B, a container vessel of 4000 tonnes, left berth 205E at the Southampton Container Terminal, and moved into the Upper Swinging Ground adjacent to the measurement position. During this period the broadband Sound Pressure Level at the monitoring position increased to a maximum of 144 db re. 1 µpa, with the perceived levels for the salmon increasing to 43 db ht (Salmo salar). The loudest levels of underwater noise recorded throughout the day occurred during the period from 16:35 to 17:07, when the container vessel Wan Hai 605 approached, and then turned in the Upper Swinging Ground before berthing at the Southampton Container Terminal. The Wan Hai 605 is a container vessel of 277.0m length and 40.0m beam. During this period the broadband Sound Pressure Level increased from approximately 120 db re. 1 µpa, to a maximum of 152 db re. 1 µpa, with the perceived levels for the salmon increasing to 53 db ht (Salmo salar). A few minutes later, the vessel Queen Victoria left berth 106, turned in the Upper Swinging Ground, and then headed up the Southampton Water shipping channel toward the Solent. During the period that the vessel was in the Upper Swinging Ground the broadband Sound Pressure Level at the monitoring position increased to a maximum of 146 db re. 1 µpa, with the perceived levels for the salmon increasing to 43 db ht (Salmo salar). The underwater noise monitoring data also indicates many transient noise peaks throughout the day. These may be attributed to container operations in the nearby container terminal, and various construction activities occurring around the Upper Swinging Ground, including ongoing work near Marchwood power station and various activities at nearby berths 107, 108 and 109. During the period from approximately 13:30 to 16:30 the noise varied as small craft including Rigid Inflatable Boats and various motor boats were active in the Upper Swinging Ground. With the exception of large shipping movements, this period had the highest continuous noise levels during the day, with broadband Sound Pressure Level at the monitoring position varying from 112 to over 130 db re. 1 µpa. The perceived levels for the salmon during this period varied from 5 to 31 db ht (Salmo salar). In summary, the background underwater noise at berth 201 / 202 varied from a broadband Sound Pressure Level of below 110 db re. 1 µpa, and a salmon perceived level of 0 db ht (Salmo 29

36 salar) during the early morning, to Sound Pressure Levels from 112 to over 130 db re. 1 µpa, and levels from 5 to 31 db ht (Salmo salar) during busy periods at the port in the afternoon. Further increases in underwater noise were recorded as both small and large vessels manoeuvred in the waters in and around the Upper Swinging Ground. Event Number Time Activity 06:15 Start of monitoring. Little harbour activity. 1 06:26 Vessel Queen Victoria approaching through shipping channel and mooring at berth :04 Queen Victoria stationary, support vessels in attendance. 3 07:25 Queen Victoria at berth 106. Lorry reversing along quay wall 201 / :17 Engines started on service barge at berth :22 10:35 Container vessel Margareta B leaves Berth 205E, moves through Upper Swinging Ground and leaves Western Docks area. 6 11:33 Fork lift truck passing along dock wall, passed measurement position. 7 12:08 Medium sized motor boat passes through Upper Swinging Ground. 8 12:51 12:57 Cargo vessel Humber Star enters Upper Swinging Ground from Bury Swinging Ground, moves past measurement position and leaves Western Docks area. 9 13:09 13:18 Gosport ferry moving through Upper Swinging Ground :00 14:42 Various small vessel movements in Upper Swinging Ground area :06 16:17 Tug boat Apex manoeuvring near Ro-Ro terminal and moving out of area :35 17:07 Large container vessel Wan Hai 605 approaching from shipping channel (Town Quay area) entering Upper Swinging Ground aided by two tugs. Vessel turns in Upper Swinging Ground and manoeuvres into Berth :06 17:34 Vessel Queen Victoria leaves Berth 106, enters Upper Swinging Ground, turns and exits Western Docks area :37 17:45 Two tugs Bentley Felixtowe and Svitzer Sussex moving through Upper Swinging Ground and moving out of Western Docks area. Table 4-2. Table of shipping movements and events in the berth 201 / 202 region during 10 th June

37 redevelopment of berths 201/202 and assessment of the disturbance to salmon. Figure Results of monitoring of underwater noise at berths 201 / 202, Southampton Container Terminal, on the 10 th June

38 5 Measurement of underwater noise from shipping. 5.1 Introduction All forms of motorised vessels generate underwater noise. The noise from large shipping is one of the dominant underwater noise sources in the sea, the review in section 4 highlighting that shipping contributes to ambient ocean noise at frequencies from 20 to 80 Hz. Individual vessels may generate very different sound levels and have different frequency characteristics depending upon factors such as the propulsion system, and whether there is propeller cavitation or singing. Richardson et al (1995) provides a review of the underwater noise from various classes of ship. For example, small ships are quoted as producing a broadband source level noise of typically 170 to 180 db re 1 m, with larger ships, such as supertankers producing underwater noise at broadband source levels of up to 190 db re 1 m. 5.2 Review of published shipping noise data Figure 5-1 presents some of the shipping noise data from Cybulski (1977), Malme et al (1989) and Richardson and Malme (1993), which are reproduced in Richardson et al (1995). The data are presented as estimated Third Octave Levels (TOLs) of source level noise, based on the extrapolation of shipping noise data measured at various far field distances. These data only present the underwater noise spectrum at low frequency, but highlight that there is considerable underwater noise energy at frequencies from 10 Hz to 1000 Hz, coinciding with the peak frequencies at which fish are able to perceive underwater noise. (It should be noted here that other ship systems such as echosounders and fish finders produce very high frequency noise that is above the frequency band shown in this data.). As part of this study these data have been re-analysed to determine source level noise for these ships in terms of the perceived level by salmon. An example of the analysis, illustrating the manner in which db ht levels for fish and other marine species are calculated, is shown in Table 5-1. In the example shown the Third Octave Levels of underwater noise for a tug / barge (column 2) have been analysed in terms of sound perception by the salmon. The hearing threshold of the salmon is shown in column 3, where the bold data are from the study of Hawkins and Johstone (1978), and the remainder of the data have been interpolated and extrapolated to cover the frequency range of the shipping data. The band sensation levels of column 4 represent the level of the sound above the hearing threshold of the salmon, in each third octave band. The data in column 4 indicates that the main components of the tug noise that contribute to the sound perception of the salmon are the third octave bands at centre frequencies of 100 Hz, 125 Hz, 160 Hz and 200 Hz. The sum of the third octave band sensation levels is the overall level of the sound above the hearing threshold of the salmon; In this case, the analysis indicates that the source level noise of the barge is at a level of 66 db ht (Salmo 1 m. Table 5-2 presents a summary of the vessel noise presented in Figure 5-1. The broadband Sound Pressure Levels are calculated from the Third Octave Levels from Richardson et al (1995). The source perceived sound levels for the salmon have been calculated using the frequency weighting process described above. The data indicate that the source levels of underwater noise perceived by the salmon, that is, the apparent levels at a nominal range of 1 m from the vessels varies from 53 db ht (Salmo 1 m for the underwater noise for the 34 m diesel vessel, to a source level noise of 87 db ht (Salmo 1 m, for the source level noise for a supertanker. These data have then been used to predict the extent of a 50 db ht (Salmo salar) zone based on spherical spreading and cylindrical spreading of the underwater noise with range. The measurements of underwater noise discussed above, however, refer to ships that are fully underway in open waters. Large vessels that pass through Southampton Water travel at much slower speeds of typically a few knots, and hence the underwater noise characteristics are likely 32

39 to be very different from the levels published in the literature. In addition, the main purpose of this study is to assess potential underwater noise sources in terms of any impact that they may have on migratory salmon. This section of the report therefore presents measurements of underwater noise from shipping passing through Southampton Water both as unweighted noise and in terms of its perception by salmon. Band centre frequency (Hz) TOL (db re 1 µpa) Salmon hearing threshold level (db re 1 µpa) Band sensation level (db) Sum of band sensation levels [db ht (Salmo salar)] 66 Table 5-1. Third Octave-band Levels (TOL s) of underwater noise estimated at a source range of 1 m from a tug pulling a barge.(richardson et al, 1995) The data are then compared with the hearing threshold of the salmon to determine band sensation levels and summed to give an overall level above hearing threshold (db ht ). 33

40 Vessel Type Source Broadband Sound Pressure Level (db re. 1 m) Source db ht (Salmo 1 m. 50 db ht (Salmo salar) range. (Spherical spreading) 50 db ht (Salmo salar) range. (Cylindrical spreading) Tug / barge m 64 m Trawler m 64 m 34 m Diesel m 2 m Supertanker m 8 km Table 5-2. Summary of source broadband (10 Hz to 10 khz) levels of underwater noise and assessment of perceived levels and impact ranges for the salmon. (Original data for shipping noise from Richardson et al, 1995). Figure 5-1. Estimated Third Octave Levels (TOLs) of underwater noise at source (at 1 m) for typical examples of shipping (Data from Richardson et al (1995), based on data from Cybulski (1977), Malme et al (1989) and Richardson and Malme (1993)). 34

41 5.3 Measurements Measurements of underwater noise from shipping in Southampton Water were undertaken during a period from March to May The underwater sound recordings were made using a Bruel and Kjaer 8106 low noise hydrophone deployed from an anti-heave buoy. The measurement equipment and procedures are described in more detail in Appendix A. The range from each vessel was measured by taking waypoints using a hand held GPS at various points along the shipping lane. The distance between the recording position and each of the waypoints was then measured as the vessel passed, and archived with the underwater sound file. Due to safety issues with large vessels that were underway, it was difficult to obtain measurements at ranges inside of a few hundred metres. It is worth noting, however, that with vessels of 100 m and greater in length, measurements at closer ranges than this would have been undertaken in the near field of the source, and so are likely to have varied inconsistently. 5.4 CMA CGM Verlaine The vessel CMA CGM Verlaine left its berth on 15 th April 2008 and passed the Dock Head, Southampton Water, at approximately 08:25. The vessel is 300 m long with a draft of 14.5 m and a deadweight tonnage of tons (See Figure 5-2). The hydrophone was deployed from the edge of the Dock Head allowing underwater sound measurements to be taken at ranges of between 220 m and 3300 m from the passing vessel. Measurements commenced at 07:30 prior to the vessel leaving berth, and continued during its transit past the Dock Head measurement position, and until it was out of site as it moved into the Solent at approximately 09:00. High Water on the 15 th April was at 07:57 and Low Water was at 13:59. Weather conditions were clear and calm with minimal wind speeds. Figure 5-3 presents a summary of the one second, Sound Pressure Levels of underwater noise with range as the vessel passed the measurement position at Dock Head. The results indicate that at a range of 220 m the broadband Sound Pressure Level (1 Hz to 120 khz) was approximately 140 db re 1μPa, decreasing to 120 db re 1μPa, at a range of 2 km. A least sum of squares fit has been applied to the data, using the underwater sound propagation model described in equation This suggests an apparent broadband source noise for the vessel at low speeds in the Southampton Water channel of 171 db re 1 m, with a sound Transmission Loss with range described by 12 log r, where r is the range in metres, and an absorption coefficient of db.m -1. It should be noted that the source level figure, and the others quoted later in this section are an estimate of the source levels of noise based on extrapolating measurements at range (i.e. in the acoustic far field) to the apparent noise level at distance of 1 m. The process of normalising the measurement to a range of one metre is the convention in acoustics (see Urick (1983), Ross (1976)). Figure 5-4 presents the results of the analysis of the underwater noise recordings in terms of the perception by the salmon. The data indicates that at a range of 220 m the noise is at a level of 37 db ht (Salmo salar). This is below the level of noise at which a behavioural disturbance would be considered likely. The perceived sound level decreases with range, until at approximately 2 km, the ship noise is at a level from 18 to 20 db ht (Salmo salar). This corresponds to the ambient sea noise on this day, and hence, as the data indicates, the level does not decrease further as the vessel moves from 2 km to a range of 3300 m. In this case the fit to the measured data indicates a source noise for the salmon of 64 db ht (Salmo 1 m, with an underwater sound Transmission Loss des described by 11 log r, and an absorption coefficient of db.m -1. Based on the extrapolation of the perceived noise for the salmon to a closer range, where the sound level increases, the data indicates a 50 db ht (Salmo salar) range of 18 m. 35

42 Figure 5-2 The CMA CGM Verlaine during exit from Southampton Docks on the 15 th April

43 Figure 5-2. Summary of one second, Sound Pressure Levels of underwater noise with range for the vessel CMA CGM Verlaine measured during exit from Southampton Docks on the 15 th April Figure 5-3. Variation of db ht (Salmo salar) underwater noise level with range from the vessel CMA CGM Verlaine measured during exit from Southampton Docks on the 15 th April

44 5.5 Kyoto Express The vessel Kyoto Express passed the Dock Head at Southampton at approximately 06:50 on the 4 th April Underwater noise measurements were undertaken as the vessel approached at approximately 06:45, passed the measurement location, and then passed through Southampton Water at 07:00 toward the Solent. The vessel Kyoto Express is shown in Figure 5-4; it is 335 m long with a draft of 12.6 m and a deadweight tonnage of 103,000 tons. High Water on the 4 th April 2008 was at 10:19 and Low Water at 16:19. Weather conditions were very calm with minimal wind speeds. Early morning fog cleared until it became bright and slightly overcast. The underwater noise measurements were undertaken with the hydrophone deployed from a small Rigid Inflatable Boat (RIB), positioned at the end of the Dock Head. Figure 5-5 presents a summary of the one second, Sound Pressure Levels of underwater noise with range as the vessel passed the measurement position at Dock Head. In this case the results indicate a variation in underwater noise depending upon whether the vessel was approaching, or moving away from the measurement position. The results with the vessel approaching indicate that at a range of 360 m the broadband Sound Pressure Level (1 Hz to 120 khz) was approximately 133 db re 1μPa, decreasing to 122 db re 1μPa, at a range of 1375 m. The least sum of squares fit to the data, suggests an apparent broadband source noise for the vessel approaching at low speeds of 169 db re 1 m, with a sound Transmission Loss with range described by 14 log r, and an absorption coefficient of db.m -1. With the vessel moving away from the measurement position the broadband Sound Pressure Level varied from 143 to 145 db re 1μPa, at a range of 120 m decreasing to approximately 125 db re 1μPa, at a range of 1500 m. The fit to the data, suggests an apparent broadband source noise for the vessel moving away at low speeds of 173 db re 1 m, with a similar Transmission Loss described by 14 log r, and an absorption coefficient of db.m -1. Figure 5-6 presents the corresponding data analysed in terms of the sound perception by the salmon. The data indicates that with the vessel approaching at a range of 360 m the noise varied from 29 to 32 db ht (Salmo salar). At a range of 1375 m the perceived noise for the salmon was at a level of approximately 14 db ht (Salmo salar). In this case the fit to the measured data indicates a source noise for the salmon of 67 db ht (Salmo 1 m, with an underwater sound Transmission Loss described by 13 log r, and an absorption coefficient of db.m -1. With the vessel moving away from the measurement position the noise level perceived by the salmon varied from 45 db ht (Salmo salar) at a range of 120 m to 18 db ht (Salmo salar) at 1500 m.. Based on the extrapolation of the perceived noise for the salmon to a closer range, where the sound level increases, the data indicates a 50 db ht (Salmo salar) range of 18 m with the vessel approaching, and a range of 70 m when the vessel is moving away. 38

45 . Figure 5-4. The Kyoto Express during exit from Southampton Docks on 4 th April

46 Figure 5-5. Summary of one second, Sound Pressure Levels of underwater noise with range for the vessel Kyoto Express measured during exit from Southampton Docks on the 4 th April Figure 5-6. Variation of db ht (Salmo salar) underwater noise level with range from the vessel Kyoto Express measured during exit from Southampton Docks on the 4 th April

47 5.6 Vega Stockholm The Vega Stockholm left berth 204, Southampton Container Terminal, at 08:25 on 4 th April Underwater noise measurements were undertaken from a RIB moored at Number 8 marker buoy at the edge of the Upper Swinging Ground. Underwater noise measurements were undertaken as the vessel moved slowly into the Upper Swinging Ground on low power, turned in the Upper Swinging Ground, and then got underway through Southampton Water to the Gymp Elbow marker buoy. Underwater sound measurements were therefore undertaken at ranges from 125 m to 2400 m. The vessel Vega Stockholm is 129 m long with a draft of 6.6 m and a deadweight tonnage of 8,307 tons (see Figure 5-7). High Water on the 4th April 2008 was at 10:19 and Low Water at 16:19. Conditions were flat calm with minimal wind speeds. The fog present during the early morning measurements cleared during the measurement period until it became brighter and slightly overcast. Figure 5-8 presents a summary of the one second, Sound Pressure Levels of underwater noise for the vessel Vega Stockholm. With the vessel moving slowly into the upper swinging ground under low power the Sound Pressure Level, measured at ranges from 275 to 300 m, varied from 133 to 142 db re 1μPa. Figure 5-9 indicates that the corresponding perceived level for the salmon during this period varied from 32 to 39 db ht (Salmo salar). With the vessel turning in the Upper Swinging Ground the Sound Pressure Level, measured at a range of 300 m (on the edge of the swinging ground), varied from 144 to 149 db re 1μPa. The corresponding perceived level for the salmon during this period varied from 39 to 47 db ht (Salmo salar). With the vessel underway through Southampton Water the broadband Sound Pressure Level varied from 153 db re 1μPa, at a range of 125 m, decreasing to 120 db re 1μPa, at a range of 2375 m, as the vessel passed the Gymp Elbow marker buoy. The fit to the measured data suggests an apparent broadband source noise for the vessel underway at low speeds through Southampton Water of 186 db re 1 m, with a sound Transmission Loss with range described by 15 log r, and an absorption coefficient of db.m -1. The analysis of the data for the salmon indicates that at ranges from 125 m to approximately 400 m, the perceived level for the salmon was fairly consistent at a level of approximately 50 db ht (Salmo salar). At a range of approximately 2 km the perceived noise had fallen to background noise level. The fit to the measured underwater noise data for the Vega Stockholm when underway indicates a source noise for the salmon of 93 db ht (Salmo 1 m, with an underwater sound Transmission Loss described by 18log r, and an absorption coefficient of db.m -1. Based on the extrapolation of the perceived noise of the Vega Stockholm when underway, where the sound level increases at closer range, the data indicates a 50 db ht (Salmo salar) range of 200 m. 41

48 Figure 5-7. The Vega Stockholm during exit from Southampton Container Terminal, berth 204, on the 4 th April

49 Figure 5-8. Summary of one second, Sound Pressure Levels of underwater noise with range for the vessel Vega Stockholm measured during exit from Southampton Docks, berth 204, on the 4 th April Data are presented for periods where the vessel is manoeuvring, then getting underway. Figure 5-9. Variation of db ht (Salmo salar) underwater noise level with range from the vessel Vega Stockholm measured during exit from Southampton Docks on the 4 th April Data are presented for periods where the vessel is manoeuvring, then getting underway. 43

50 5.7 Red Jet Ferry The Red Funnel, Red Jet ferries run at regular intervals between Town Quay, Southampton, and the Isle of Wight (see Figure 5-10). Between the times of 15:20 and 15:40 on the 7 th May 2008, two separate Red Funnel ferries, passed the underwater noise measurement position opposite Netley Abbey, during transit through Southampton Water. The vessels were moving at speeds in excess of 10 knots during their transit to and from the Isle of Wight. Underwater noise measurements were made from a hydrophone deployed from a RIB. The measurements commenced prior to the ferries reaching the monitoring position, and continued until after the underwater noise from the vessels had fallen below background levels. Underwater noise recordings were undertaken at ranges from 200 m, to a range of 1500 m from the moving vessels. High Water on the 7 th May was at 12:43 with Low Water at 18:43. Weather conditions were clear and sunny with a moderate South-westerly breeze and slight chop on the water. Figure 5-11 presents a summary of the one second, Sound Pressure Levels of underwater noise with range as the first vessel (Run 1) passed the measurement position at Netley Abbey. The measurements were undertaken with the Red Jet ferry at ranges from 100 m to 1000 m. The results indicate a relatively low level of underwater noise compared that for the larger container vessels moving through Southampton Water. At a range of 100 m the data indicates a broadband, Sound Pressure Level (1 Hz to 120 khz), of approximately 131 db re 1μPa, decreasing to 124 db re 1μPa, at a range of 1000 m. The fit to the data suggests an apparent broadband source noise for the vessel of 154 db re 1 m, with a sound Transmission Loss with range described by 10 log r. Figure 5-12 presents the results in terms of the perception by the salmon. The data indicates that at a range of 100 m the noise is at a level of 24 db ht (Salmo salar). This is below the level of noise at which a behavioural disturbance would be considered likely. The perceived sound level decreases with range, until at 1000 m, the vessel noise is at a level from 12 to 14 db ht (Salmo salar). This corresponds to levels at ambient sea noise on this day. The fit to the measured data indicates a source noise for the salmon of 56 db ht (Salmo 1 m, with an underwater sound Transmission Loss described by 14 log r, and an absorption coefficient of db.m -1. Figures 5-13 and 5-14 present the unweighted Sound Pressure Levels and perceived levels of noise for the salmon measured as the second Red Jet ferry passed the monitoring point. For this case the data indicates a similar level of noise with range, giving an apparent broadband source noise for the vessel of 155 db re 1 m. The perceived sound levels varied from 23 db ht (Salmo salar) at a range of 200 m, to sea noise levels from 13 to 18 db ht (Salmo salar) at ranges of approximately 1 km. Based on the extrapolation of the perceived noise for the salmon to a closer range, where the sound level increases, the data indicates a 50 db ht (Salmo salar) range of no more than a few metres. 44

51 Figure The Red Funnel, jet ferry, Red Jet 3 entering Town Quay, Southampton Water on the 7 th May,

52 Background Figure Summary of one second, Sound Pressure Levels of underwater noise with range for a Red Funnel, jet ferry, passing Netley Abbey, Southampton Water, on the 7 th May, 2008 (Run 1). Background Figure Variation of db ht (Salmo salar) underwater noise level with range from a Red Jet, jet ferry passing Netley Abbey, Southampton Water, on the 7 th May, 2008 (Run 1). 46

53 Figure Summary of one second, Sound Pressure Levels of underwater noise with range for a Red Funnel, jet ferry, passing Netley Abbey, Southampton Water, on the 7 th May, 2008 (Run 2). Background Figure Variation of db ht (Salmo salar) underwater noise level with range from a Red Jet, jet ferry passing Netley Abbey, Southampton Water, on the 7 th May, 2008 (Run 2). 47

54 5.8 Summary of shipping noise impact on salmon Table 5-3 provides a summary of the measurements of underwater noise for typical vessel traffic moving through Southampton Water. The larger container vessels were moving at considerably slower speed than similar vessels in the open seas. The levels are relatively low and it is concluded that shipping activities do not adversely affect salmon behaviour. Vessel Vessel Type Source Broadband Sound Pressure Level (db re. 1 m) Source level db ht (Salmo 1 m. 50 db ht (Salmo salar) range. CMA CGM Verlaine Kyoto Express Approaching Kyoto Express Moving away Vega Stockholm Red Jet Ferry Run 1 Red Jet Ferry Run m draft container vessel 12.6 m draft container vessel 12.6 m draft container vessel 6.6 m draft container vessel Jet hydrofoil > 10 knots. Jet hydrofoil > 10 knots m m m m m Table 5-3.Summary of broadband (1 Hz to 120 khz) source Sound Pressure Levels and db ht (Salmo salar) levels of underwater noise from measurements of shipping noise in Southampton Water, March to May Based on the underwater noise measurements, estimates are also provided of the range for the shipping noise to fall below 50 db ht (Salmo salar) 48

55 6 Underwater noise from construction operations during berth 201/202 development. 6.1 Introduction The Scoping Study for the Southampton 201 / 202 berth development highlighted the need for an appropriate assessment of the underwater noise impact of the proposed construction activities. In particular, the assessment must consider the impact of underwater noise from the construction activities on the Annex II protected species, the salmon (Salmo salar). A review of the proposed construct undertaken with Associated British Ports and Jacobs identified that the main construction activities generating noise in the underwater environment are; Impact pile driving operations to secure the toe of each pile of the new quay wall into the hard underlying substrate. Above water rock breaking operations to break out the concrete from the existing quay wall at berths 201 / 202. Underwater rock breaking operations to break out the existing quay wall below the water line This section of the report provides an assessment of the underwater noise from these activities. 6.2 Piling operations There are many types of pile, but the most commonly used sorts are tubular piles, mainly used for foundations, and sheet piles, which link together. These may be used temporarily to form walls for excavations, or permanently, for instance as retaining walls for docks. The most commonly used method for securing the pile is impact piling. This technique can generate high levels of airborne and underwater noise. Hence, over recent years many attempts have been made to reduce the airborne noise by developing alternative piling techniques and equipment. The techniques that can be used to secure piles into a sea or riverbed are briefly reviewed below. However, due to the compact nature of the ground at the berth 201 / 202 site, it is considered unlikely that techniques other than impact piling would be effective. The subsequent review of underwater noise from the berth development therefore concentrates on the use of percussive impact piling Vibro-piling The principle of operation of a vibropiling system is that counter-rotating, out-of-balance masses that are geared together rotate in an enclosure attached to the top of the pile. The rotating masses generate a resultant vertical vibratory force that slowly forces the pile into the ground. The vibropiling technique typically generates lower levels of noise and vibration than percussive impact piling. The technique works well in soft substrates, but may not be effective in hard ground Hydraulic piling Hydraulic piling is a method used to install and extract steel piles with the use of static-load hydraulic pressure. The hydraulic piling machine is initially attached to a reaction stand in order to obtain a reaction force to drive the first piles. Alternatively, the initial piles may be impact driven. The hydraulic piling machine is then attached to the secure piles, using them as a base to drive the next sheet pile. The machine then slides forward and uses the resistance of the previously installed piles to drive (push in) the next pile to be driven. This method of pile installation has not yet been developed for the required dimension of the piles, combined with the hard ground conditions, and is considered unfeasible for this development. 49

56 6.2.3 Bored (Auger) piling There are a number of variants of auger bored piling techniques developed to overcome access and environmental restrictions. The techniques involve an auger string being rotated into the ground until the desired depth is reached. In some cases the pile may be pulled down behind the auger, or the auger string may be withdrawn and the pile introduced and either concreted or grouted to secure the pile. The technique has the advantage of producing relatively low noise and vibration, but has limited ability to penetrate significant obstructions or bedrock Impact Piling The impact piling technique involves a large weight or ram being dropped or driven onto the top of the pile, driving it into the ground. Usually, double-acting hammers are used in which compressed air not only lifts the ram, but also impart a downward force on the ram, exerting a larger force than would be the case if it were only dropped under the action of gravity. Percussive impact piling has been established as a high level source of underwater impulsive noise (Wursig 1993, Caltrans 2001, Nedwell et al 2003, Parvin et al 2006, Thomsen et al, 2006, Nedwell et al 2007). Noise is created in air by the hammer, partly as a direct result of the impact of the hammer with the pile. Some of this airborne noise is transmitted into the water. Of more significance to the underwater noise, however, is the radiation of noise from the surface of the pile as a consequence of the compressional, flexural or other complex structural waves that travel down the pile following the impact of the hammer on its head. As water is of similar density to steel and, in addition, due to its high sound speed (1500 m.s -1, as opposed to 340 m.s -1 for air), waves in the submerged section of the pile couple sound efficiently into the surrounding water. These waterborne waves will radiate outwards, usually providing the greatest contribution to the underwater noise. At the end of the pile force is exerted on the substrate not only by the mean force transmitted from the hammer by the pile, but also by the structural waves travelling down the pile inducing lateral waves in the seabed. These may travel as both compressional waves, in a similar manner to the sound in the water, or as a seismic wave, where the displacement travels as Rayleigh waves. The waves can travel outwards through the seabed, or by reflection from deeper sediments. As they propagate, sound will tend to leak upwards into the water, contributing to the waterborne wave. Since the speed of sound is generally greater in consolidated sediments than in water, these waves usually arrive first as a precursor to the waterborne wave. Impact piling is the proposed method of pile installation given the required size of piles (1800mm diameter) and the hard ground conditions. 6.3 Underwater noise from pile driving operations As part of this study, the propagation of underwater noise from the proposed pile driving operations has been modelled on a depth-dependant basis, in order to provide an estimate of underwater sound level at range from source. These data have been used to predict the Source Level noise and underwater sound propagation parameters associated with the berth 201 / 202 development. Transmission of sound in the underwater environment is highly variable from region to region, and can also vary considerably with the local bathymetry. In particular, the low frequency components of piling noise can be rapidly attenuated in very shallow water regions typical of estuaries, and the silt and sandbanks regions located around the UK coast. In general, for deep water propagation, the lower the frequency of sound, the more easily the sound propagates. High frequency components, by contrast, are more easily attenuated underwater. Optimal sound propagation conditions are met when the water depth is approximately uniform or increasing. By contrast, in shallow water, or when the water depth decreases with range, there is a greater interaction of the sound with the seabed, and the sound is more easily absorbed. 50

57 In addition to the frequency spectrum of the outgoing signal and the local bathymetry, the nature of the seabed sediments may also influence sound propagation. In general, the softer and more unconsolidated the seabed sediments, the greater the attenuation of sound with distance and the lower the Sound Pressure Levels become. Seabeds consisting of soft mud or silt sediments generally produce greater propagation losses than those consisting of harder sediments such as rock and shingle. In very shallow waters, typical of estuaries and shoals, the underwater sound will vary randomly both temporally and spatially due to many factors. The approach that Subacoustech have used in many previous studies is to base the modelling and assessment on measured data in similar circumstances. This data indicates that there is a greater interaction of underwater sound with the seabed in shallow coastal areas than in deep water, resulting in propagation losses which typically increase with increasing frequency, but decrease with increasing depth. The Impulse Noise Sound Propagation and Impact Range Estimator (INSPIRE) model has been developed by Subacoustech on this basis. It uses a combined geometric and energy flow/hysteresis loss model to model propagation in shallow water. The INSPIRE data for the deeper water sites, of approximately 15 m and greater, is based on, and has been tested against, measurements during impact piling to construct the Barrow offshore wind farm (Parvin et al. 2006). For propagation in very shallow water coastal regions the model has been tested against the measured data reported for piling operations during the Kentish Flats offshore wind farm development (Parvin et al, London Array) The INSPIRE model has also been tested against impact piling data from the North Hoyle, Scroby Sands (Nedwell et al, 2007) and Beatrice offshore wind farm construction operations, as well as range of shallow water estuarine piling operations Characteristics of underwater noise from pile driving operations Many of the features of the underwater noise propagated during impact piling operations are similar to that for an underwater blast wave. At close range impact piling noise is typically characterised by a transient pressure wave, reaching its peak pressure level in approximately 1 ms, with reflections, reverberation and ringing of the pile extending the total time history for each impact event to a duration of several hundred milliseconds. At greater range the transient event is further spread and can last for up to a second. The pile is usually struck at a repetition rate of one strike every 1 to 2 seconds, for a period of several hours to secure the pile, depending upon the substrate and the pile penetration depth required. As an example, Figure 6-1 presents an underwater noise time history recorded at a range of 76 m from sheet piling operations at the Marchwood power station site, Southampton Water. These measurements were undertaken by on behalf of the Environment Agency. The individual pile strikes are clearly characterised by the sharp peaks in pressure level. The individual pile strikes vary from approximately 500 Pa to a maximum peak to peak level of 1440 Pa or 183 db re 1 μpa. The broadband RMS level of the sound over this period was 149 db re. 1 μpa. Figure 6-2 presents a typical noise time history for measurements taken further away from the piling works at a range of 225 m. At this range the piling noise reached a maximum peak to peak level of 106 Pa or 161 db re 1 μpa. The RMS sound level here was measured at 145 db re. 1 μpa. The underlying low frequency variation in pressure evident in Figure 6-2 is due to wave motion causing hydrostatic pressure changes. For comparison, Figure 6-3 illustrates a similar underwater noise time history recorded at a range of 92 m from an offshore pile driving operation to secure a 4.7 m diameter tubular steel pile into the seabed (Parvin et al, London Array). In this case, the data indicates that the transient pressure wave has a peak to peak level of 70,000 Pa, or 217 db re. 1 μpa. At this range the initial peak of the waveform is very pronounced, and is of very short duration, lasting for approximately 1 ms, whilst the pressure waveform has a total duration of approximately 20 ms. 51

58 Figures 6-4 presents an underwater noise time history data measured at distances of 12.8 km from the same impact piling operation. At this range, the transient pressure wave from the impact piling operation is still clearly distinguishable, with a peak to peak sound level of approximately 450 Pa, or 173 db re. 1 μpa. The total duration of the pressure wave for each pile strike at this range extends to approximately 300 ms, hence, although the magnitude of the pressure wave has decreased with range, the wave extends over a considerably greater time period. Figure 6-5 provides a comparison of the spectral content of the transient pressure waves from impact piling measured at ranges of 92 m and 12.8 km from the impact piling operation with 4.7 m diameter tubular steel piles. At a range of 92 m, the one second averaged noise spectrum (Power Spectral Density) indicates that the peak underwater sound energy from piling occurs over the frequency range from 70 Hz to 500 Hz, where the spectral levels vary from approximately 150 to a maximum of 170 db re. 1 μpa 2.Hz -1. Spectral levels from 500 Hz to 100 khz vary from 150 to 100 db re. 1 μpa 2.Hz -1 indicating that at short range the piling noise is very broadband, and there is considerable high frequency content in the pressure wave. At a range of 12.8 km from the same impact piling operation, the high frequency components of the piling noise have decreased rapidly, and the underwater noise is dominated by low frequency components of sound Source Level noise from impact piling Over the past five years, have undertaken a number of studies, and comprehensive reviews of underwater noise during impact piling operations for offshore wind farm development (Nedwell et al (2003), Nedwell et al (2007), Parvin et al (2006b)). The data for piling operations in relation to the construction of offshore wind farms is summarised in Table 6-1. It is based on the data of Nedwell et al (2003) and Parvin et al (2006b) for the North Hoyle and Scroby Sands offshore wind farm sites, the data reported by Parvin at al (2006b) for the Barrow and Kentish Flats offshore wind farm construction operations, and the data of Parvin et al (2006d) for the Burbo Bank offshore wind farm construction. Offshore wind farm Peak-to-peak Source Level Pile diameter Water depth development (db re. 1 1 m) Barrow 4.7 m 10 to 20 m 252 North Hoyle 4.0 m 10 to 15 m 249 Scroby Sands 4.2 m 3 to 30 m 257 Kentish Flats 4.3 m 5 to 8 m 243 Burbo Bank 4.7 m 15 m 250 Table 6-1. Summary of Source Level impact piling noise from measurements during offshore wind farm construction in UK waters. Studies undertaken for the Crown Estate and the Environment Agency, have measured the underwater noise from smaller scale impact piling operations in rivers and estuaries. Source Levels for 0.5 to 0.9 m diameter tubular pile, impact piling operations have been estimated from the data of Nedwell et al (2002a, 2002b and 2003b). These indicate a noise Source Level of 189 db re. 1 1 m for a 0.5 m diameter pile in very shallow water, 211 db re. 1 1 m for a 0.7 m diameter pile, and 201 db re. 1 1 m for a 0.9 m diameter pile. The underwater noise levels from 2.4 m diameter tubular steel piling operations are available from Abbott et al (2002). Re-analysis of this data indicates a Source Level noise of 242 db re. 1 1 m. 52

59 Further recent data is also available from the data of Bailey et al (in preparation). This indicates a Source level noise of 234 db re. 1 1 m for 1.8 m diameter tubular steel pile driving operations in relatively deep water (42 m), during construction of the Beatrice offshore wind farm development project. These studies have indicated that the apparent source level noise from percussive pile driving operations is related to the dimensions of the pile (diameter or width), the water depth, and the blow force required to achieve pile penetration through the bed substrate. Data from previous impact piling operations in similar depth waters indicates (15 m) indicates a noise Source Level for 1800 mm and 1000 m tubular steel piles and 600 mm width sheet piles during the berth 201 / 202 development as indicated in Table 6-2. It is understood that the requirement for driven piles is based on the densely compacted sediments and substrates at the site. It is likely that the piles will have to be driven into the underlying bedrock, and hence a hard substrate has been assumed in the Source Level estimates and sound propagation modelling. Pile Type Peak to peak Source noise (db re. 1 1 m) Peak to peak noise dbht 1 m 1800 mm tubular steel mm tubular steel mm sheet pile Table 6-2. Predicted peak-to-peak Source Level noise for proposed pile driving operations at the Southampton Container Terminal, berth 201 / 202 site. 6.4 Sound propagation modelling Introduction The acoustic propagation modelling is based on four transects from the site (See Figure 6-6). These have been selected to illustrate the shortest and longest range propagation paths from the construction site. The modelling has considered the underwater noise during 1800 mm and 1000 mm diameter tubular steel pile, impact pile driving operations, and the underwater noise from 600 mm width sheet pile driving operations. No allowance has been made in the modelling for the contribution to the overall noise that may occur from reflections from the existing quay wall immediately behind the piling operations. As the pile driving operations are being conducted close to the existing quay wall in principle some augmentation of the source might be expected. It is not thought that this would be significant. The proposed site is at the edge of the upper swinging ground, at the upper end of the Southampton Water shipping channel. The site is in relatively deep water, varying from approximately 9 to 11 m at Lowest Astronomic Tide (LAT). The water depth around the site remains fairly constant, until at a range of 200 m the water depth decreases to shallower regions toward the southern side of the Eling Channel or the River Test. Table 4-2 provides a summary of the tidal range (water depth above LAT chart datum at neap and spring tides) for Southampton Water (Admiralty chart ). At Mean High Water Springs (MHWS) the water depth is typically 4.5 m above LAT, and at Mean Low Water Springs (MLWS) the Admiralty chart data indicate a water depth of 0.5 m above LAT. For the purposes of the acoustic modelling undertaken in this study a mean water depth of 3 m above LAT chart datum has been used to represent the water depth at the site and along the sound propagation transects. The modelling therefore represents underwater sound propagation at mid-tide. 53

60 Latitude Longitude Height in metres above datum Place N W MHWS MHWN MLWN MLWS Southampton 52 o 54 1 o Table 6-3.Summary of tidal range data for Southampton Water (Admiralty Chart ) mm diameter tubular steel piling Figure 6-8 presents the results of modelling of the underwater noise with range for 1800 mm diameter tubular steel piles for the berth 201 / 202 development. The data are presented in the form of unweighted peak to peak sound levels. As the water depth is consistent across the upper swinging ground and River Test Channel, each of the sound propagation transects indicates a similar level of noise for the initial 250 m. Over this range the unweighted peak to peak noise level is predicted to decrease from an initial Source Level of 234 db re. 1 1 m, to approximately 188 db re. 1 μpa at a range of 250 m. Beyond this range there is some variation in the predicted noise depending upon the water depth profile. For sound propagation along the River Test channel the noise level is predicted to remain at an unweighted peak to peak level above 170 db re. 1 μpa to a range of approximately 2 km. Section 3 of this report reviewed the levels of unweighted noise that may cause lethality and physical injury to fish and other marine animals. The source level predicted for the pile driving operations is below levels likely to cause lethality to fish, but above the level at which physical injury might occur. Based on a received level of 220 db re. 1 μpa, physical injury to fish might occur to a range of approximately 6 m. If the more conservative criteria for fish as small as 0.01g is used (Popper et al, 2006), then this would extend this range to approximately 10 m. Figure 6-9 presents the predicted peak to peak db ht level with range for the salmon, for each of the acoustic propagation transects considered. The data indicates that the predicted noise levels for the salmon decrease from an initial Source Level of 134 db ht (Salmo 1 m, to 87 db ht (Salmo salar) at 250 m. The data indicates that the perceived level of noise for the salmon remains above 50 db ht (Salmo salar) for the extent of the sound propagation transects across Southampton water to Millbrook and Marchwood mm tubular steel piling Figure 6-10 presents the corresponding salmon noise data predicted for 1000 mm diameter tubular steel pile driving operations at the berth 201 / 202 development site. The data shows a similar trend to that for the larger piles, but with a reduced source noise level as indicated in Table 6-2. For this case the data indicates that the perceived noise is likely to remain above 90 db ht (Salmo salar), and therefore to cause a strong behavioural avoidance response in salmon to a range of 60 m. The perceived sound level is predicted to remain above 50 db ht (Salmo salar), and may therefore cause a behavioural disturbance throughout the river crossing area in the vicinity of berths 201 / mm width sheet piling Figure 6-11 presents the predicted salmon noise data for 600 mm width sheet piling at the berth 201 / 202 development site. For this case the data indicates that the perceived noise is likely to remain above 90 db ht (Salmo salar), and therefore to cause a strong behavioural avoidance response in salmon to a range of 15 m. Although predicted to be at a lower perceived sound level than the noise from the larger dimension tubular steel pile driving operations, the underwater noise is predicted to be above 50 db ht (Salmo salar), and may therefore cause a behavioural disturbance throughout the river crossing area in the vicinity of berths 201 /

61 6.5 Underwater noise from concrete breaking on the existing quay Tools that impact with a firm substrate or structure generate high levels of noise and vibration. As the acoustic impedance of water and solid structures are broadly similar, the vibrations can couple well into surrounding water courses, where high levels of underwater noise can result. Figure 6-12 presents the only known measurement of the underwater noise time history recorded at a distance of approximately 400 m from operations with a rock pecker on a dockside at the River Tees in The rock pecker was being used to break-up the concrete dockside to create a channel for fitting cabling. The rock pecker was approximately 50 m from the shore, with the survey boat and hydrophone in the centre of the water channel at a further distance of 300 to 400 m. The rock pecker operated for the initial 2.5 seconds of the sound file. The remainder of the recording indicates the background underwater noise in the region. The unweighted noise data indicates a rapid increase in waterborne noise, each time that the rock pecker hammers against the concrete dockside. At a range of approximately 400 m the data indicates a peak to peak noise level of 100 Pa (160 db re. 1µPa). Figure 6-13 presents the analysis of the same underwater noise recording. The data highlights that although the noise is clearly detectable using the underwater sound recording equipment, the analysis indicates that the levels are below 50 db ht (Salmo salar) when processed for perception by the salmon. Similar analysis is presented in Figures 6-14 and 6-15 during a continuous 5 second period of rock pecker concrete breaking operations. In this case the mean peak to peak sound levels were measured at 43 db ht (Salmo salar). The underwater noise from rock pecker operations above the water line on the quayside might be expected to exceed a peak to peak level of 50 db ht (Salmo salar) within a range of 200m. It should be noted that this measurement was made on an opportunity basis when rock breaking was serendipitously being undertaken in the vicinity of measurements which were being made of other noise sources. Consequently, the authors did not have access to the site on which the breaking was occurring, or contact with the staff using the breaker. Consequently, the details of the machine being used, the material being broken and the exact distances involved are not known. As there are no other measurements available, the above data are being included in this report as the best information currently available, and as such should be treated with caution. 6.6 Underwater noise from rock breaking operations below the water line At present there are no data available for rock pecker operations where the impact head is below the waterline. 6.7 Summary of construction noise impacts on salmon The modelling of underwater noise from percussive and steel sheet pile driving operations during the berth 201 / 202 development indicates that the levels are likely to exceed 50 db ht (Salmo salar) at the quay wall, and in the waters to the adjacent shoreline at Marchwood. The underwater noise may therefore cause a disturbance to migratory salmon. For operations close to the dockside, or at ranges inside of 200 m, the underwater noise from rock pecker operations on the quayside may exceed a peak to peak level of 50 db ht (Salmo salar). Rock pecker operations conducted below the waterline would be expected to generate higher levels of noise, although at present there are no data available for these forms of operation. 55

62 Figure 6-1. An underwater noise time history at a range of 76 m from sheet piling operations at the Marchwood power station site ( ). Figure 6-2. An underwater noise time history at a range of 225 m from sheet piling operations at the Marchwood power station site ( ). 56

63 Figure 6-3. A 5 second time history of underwater noise measured at a distance of 92 m from impact piling during construction of the Barrow offshore wind farm (Parvin et al, 2006). Figure 6-4. A 5 second time history of underwater noise measured at a distance of 12.8 km from impact piling during construction of the Barrow offshore wind farm (Parvin et al, 2006). 57

64 Figure 6-5. Comparison of spectral levels for subsea impact piling noise measured at ranges of 92 m and 12.8 km, during construction of the Barrow offshore wind farm (4.7 m diameter piles). (Parvin et al, 2006). Figure 6-6. Location of proposed construction works and illustration of underwater sound propagation transects selected. 58

65 Figure 6-7. Variation of water depth along underwater sound propagation transects selected. Figure 6-8. Predicted unweighted peak to peak sound level with range for 1800 mm tubular steel pile driving operations at berths 201 /

66 Figure 6-9. Predicted peak to peak db ht (Salmo salar) with range for 1800 mm tubular steel pile driving operations at berths 201 / 202. Figure Predicted peak to peak db ht (Salmo salar) with range for 1000 mm tubular steel pile driving operations at berths 201 /

67 Figure Predicted peak to peak db ht (Salmo salar) with range for 600 mm width sheet pile driving operations at berths 201 /

68 Figure An underwater noise time history during rock pecker construction operations at a range of approximately 400 m [File: 15 Nov, ]. Figure Variation of Sound Pressure Level and db ht (Salmo salar) during rock pecker construction operations at a range of approximately 400 m [File: 15 Nov, ]. 62

69 Figure An underwater noise time history during rock pecker construction operations at a range of approximately 400 m [File: 15 Nov, ]. Figure Variation of Sound Pressure Level and db ht (Salmo salar) during rock pecker construction operations at a range of approximately 400 m [File: 15 Nov, ]. 63

70 7 Underwater noise from channel dredging operations. 7.1 Introduction This section of the report presents measurements and analysis of the underwater noise during suction dredging and backhoe (back-excavation) dredging operations. There are broadly two forms of dredging operation depending upon the method used to transport the loosened material from the seabed to the water surface, and into the supporting vessel or cargo hold. These can be described as hydraulic (suction) dredgers or mechanical dredgers. 7.2 Mechanical dredging Introduction Mechanical dredging operations involve the use of some form of grab or bucket to loosen the seabed material, and then to raise and transport the material to the sea surface. There are several techniques. A Bucket dredger has a continual chain of buckets that fill by scrapping over the seabed and then empty by turning upside down by passing over a tumbler at the top. A Grab dredger has a large mechanical grab consisting of two half shells that are used to pick up material from the seabed, and then to lift and place the excavated material into a support barge. A Backhoe dredger is a mechanical excavator equipped with a half-open bucket. The excavator is filled by moving the bucket toward the machine, scrapping it along the seabed. The dredged material is then lifted to the surface in the bucket, where it is normally loaded into a support barge Underwater noise measurements during backhoe dredging operations A series of underwater noise measurements were undertaken on 29 th May 2008 during harbour dredging and reclamation works at Lerwick, Shetland. Part of this works involved the use of a backhoe dredger, the Manu Pekka. The dredger was of typical construction and it was thought that the levels of noise would be similar to those for the proposed works. The dredging operation involved the removal of material from the seabed by means of a large back excavator positioned at the rear of the Manu Pekka platform. The dredged material was then deposited into a cargo vessel moored beside the Manu Pekka. The cargo vessel then transported the dredged material to a reclamation area at the nearby Greenhead Base. On the 29 th May 2008, the Manu Pekka was positioned in the North Harbour, Lerwick, in an area at approximately the middle of the channel between Gremista in the west and Heogan to the east. All underwater noise measurements were recorded by means of a Bruel & Kjaer Type 8106 hydrophone together with associated signal conditioning and power amplifiers (See Appendix A). The equipment was deployed either from the Manu Pekka for short range measurements, or the survey vessel Alluvian for measurements at longer range. The water depth in the survey area varied from approximately 6 m near the dredging operation to 30 m in the area to the north of the harbour. Weather conditions throughout the measurement period were breezy and overcast, with some sunny intervals. Wind speeds varied from 4.5 m/s to 5 m/s. Sea conditions were choppy with frequent white horses. Due to the sheltered position there was very little swell in the survey area Results of underwater noise measurements during backhoe dredging Figure 7-1 presents an underwater noise time history with the recording hydrophone at a distance of 7 m from the backhoe dredging operation. In general, the noise is characterised by the underwater noise from ship generators and movement of the mechanical arm of the excavator bucket. The noise during these periods is clearly audible on the recordings, and is typically at a peak to peak level of 100 Pa, with one second, RMS sound pressures for these 64

71 periods of approximately 10 to 15 Pa (140 to 145 db re. 1µPa). The underwater noise during these periods is illustrated from 0 to 10 seconds of the file illustrated in Figure 7-1. When the bucket strikes the seabed there is an increase in noise, followed by further higher levels of underwater noise as the bucket lift is scrapped along the seabed. A typical period of underwater noise during this phase of the dredging operation can be seen during the period from 10 to 27 seconds of the recording in Figure 7-1. During this period the underwater noise increased to a maximum peak to peak level of 900 Pa. The one second, RMS sound pressures during this period varied from 15 Pa to 130 Pa (145 to 162 db re. 1µPa). At 7 m range the loading of the spoil into the supporting barge can be heard on the underwater noise recordings, but this activity appears to add little to overall noise. Figure 7-2 illustrates a similar underwater noise time history recorded at a range of 105 m from the Manu Pekka during backhoe dredging operations. At this range the mechanical bucket scrapping along the seabed is still clearly audible on the underwater sound recordings, and can be seen to increase the underwater noise during the periods from 13 to 23 seconds and 27 to 36 seconds in the recording shown. At 105 m range the underwater noise reached maximum peak to peak levels of approximately 120 Pa, with one second RMS levels of underwater noise reaching a maximum level of 18 Pa, corresponding to a Sound Pressure Level of 145 db re. 1 µpa. The general underwater noise from the backhoe dredging operation at this range was at a peak to peak level of approximately 30 Pa, with one second RMS levels of underwater noise from 4 Pa (132 db re. 1µPa) to 6 Pa (136 db re. 1µPa). Figure 7-3 presents an underwater noise time history recorded at a range of 2.2 km from the Manu Pekka during backhoe dredging operations. At this range the dredging noise could not be heard on the underwater sound recordings, and did not increase the noise above the periods of ambient sea noise when no dredging was taking place. The small variations illustrated in the underwater noise time history are due to hydrostatic pressure changes as waves pass over the hydrophone. The data indicates that the ambient sea noise was at a peak to peak level of approximately 1 to 5 Pa, giving RMS sound pressures from 0.1 (100 db re. 1µPa) to 1 Pa (120 db re. 1µPa). Figure 7-4 presents the spectral levels of underwater noise at increasing range from the backhoe dredger operations. In the immediate vicinity of the dredger the data indicates high levels of very low frequency noise over the frequency range from 1 Hz to 40 Hz. This noise appears to decay rapidly, with the data measured at a range of 105 m indicating considerably lower spectral levels of noise over this frequency range. The spectral levels of noise from 20 Hz to 10 khz indicate a consistent decrease in level with range from the dredging vessel. Figure 7-5 provides a comparison of the spectral levels of underwater noise during each phase of the dredging operation (i.e. Movement of the arm from the support barge to the water, dredging of material from the seabed, and movement of dredged material from the seabed and depositing into the support barge.). These data were obtained from an underwater noise time history at a range of 215 m. The data indicates that the dredging activity increases the underwater noise at frequencies from 20 Hz to approximately 20 khz. The data indicates a consistent noise over the low frequency range from 20 to 80 Hz, with peak spectral levels of noise from 35 to 45 Hz. This noise is present regardless of the phase of the dredging arm activity, and hence suggests that this noise is most likely a feature of the ship or dredger generator system(s). Similar spectral levels of underwater noise occur during movement of the bucket dredger arm from the support barge to the water surface, and when the dredged material is lifted from the seabed and place into the receiving vessel. During these operations the measurements at a range of 215 m indicate that the spectral levels of underwater noise were at a level of 80 db re. 1 µpa 2.Hz -1, over the frequency range from 80 Hz to 1000 Hz, falling to 40 db re. 1µPa 2.Hz -1, over the frequency range from 1000 Hz to approximately 20 khz. With the bucket in contact with the seabed, and material being excavated, the spectral levels of noise presented in Figure 7-5 indicate an increase in underwater noise over the frequency range from approximately 65

72 150 Hz to 20 khz. Hence, during the extraction of material from the seabed by the backhoe dredger, the noise increased to a spectral level of 85 to 100 db re. 1µPa 2.Hz -1, over the frequency range from 80 to 1000 Hz. The impact of the backhoe dredger bucket and subsequent impacts with dredged material can clearly be heard on the underwater sound recordings. Figure 7-6 presents a summary of one second Sound Pressure Levels of underwater noise measured during backhoe dredging operations at Lerwick. The figure presents the results of 1850 individual one second data samples, recorded at ranges from 7 m to 1500 m from the backhoe dredger. The data is also summarised in Table 7-1. There is considerable variability in the data at each measurement position, due to the changes in the radiated noise throughout the backhoe dredging process. At a range of 7 m, for example, the one second Sound Pressure Levels vary from 131 to 162 db re. 1µPa. The higher levels correspond to periods when the bucket excavator was in contact with the seabed. Table 7-1 illustrates that the mean Sound Pressure Level during the backhoe dredging operations varied from 143 db re. 1µPa at a range of 7 m, to background sea noise levels at typically 110 to 120 db re. 1µPa, at ranges beyond approximately 1 km. Two forms of fit to the measured underwater noise data from the dredging operation are presented in Figure 7-6., each using the underwater sound propagation model of equation The lower sound level fit indicates the least sum of squared fit to the measured data, incorporating all of the 1850 one second samples of underwater noise. This indicates an apparent Source Level noise of 154 db re. 1 m, with the underwater sound decreasing with range described by the 11 log r (where r is the range in metres), and with an absorption coefficient of db.m -1. However, this fit to all of the data is largely based on periods of underwater noise recording when the backhoe dredger bucket is not in contact with the seabed, and hence describes the general level of the vessel and moving backhoe dredger arm noise. If only the higher levels of underwater noise are considered at each measurement position, corresponding to periods when the material was being excavated from the seabed, then the second higher fit to the data is obtained. This indicates a Source Level noise of 163 db re. 1 m, with the underwater sound decreasing with range described by the 10 log r. All of the measurements of underwater noise during backhoe dredging operations at ranges from 7 m to 2200 m, and the prediction of the apparent Source Level noise of the operation at a range of one metre, are below the levels of broadband underwater noise likely to cause fatality or physical injury to species of fish Behavioural avoidance The most important factor in determining the extent of any behavioural impact zone for salmon from the dredging operation is to determine the amount of sound energy above the auditory threshold of the species. Figure 7-7 presents the results of the analysis of the underwater noise time history data in terms of its sound perception (loudness) for the salmon. These data have been calculated by passing the high frequency underwater sound recordings through time domain filters that mimic the auditory response of salmon, in the same way that airborne sound is filtered for human sound perception using A-weighting (A full description of the weighting process is provided in section 3.4). The variability of these data at each measurement position is summarised in columns 4, 5 and 6 of Table 7-1. Of the 279, one second data samples measured at a range of 7 m, the RMS perceived level for the salmon varied from a maximum of 58 db ht (Salmo salar) to a minimum of 25 db ht (Salmo salar), with a mean level of 37 db ht (Salmo salar). These perceived sound levels are considerably below those at which a strong behavioural avoidance to sound might occur. In human terms, sound levels of 50 to 60 db above hearing threshold are typical of those of an office environment. At the 50 m measurement position all of the measured db ht (Salmo salar) levels were below the 50 db ht level often used as a conservative behavioural impact criteria for migratory fish. 66

73 Figure 7-7 presents the perceived levels of backhoe dredging noise with range for the salmon. The data indicates a gradual decrease in level with range, but also the same variability of results at each measurement position due to the changes in radiated noise throughout the phases of the backhoe dredge. The least sum of squares fit to all of the measured data indicates an apparent Source Level noise of 48 db ht (Salmo 1 m, with a propagation loss described by 12 log r, and an absorption co-efficient of db.m -1. Even at a source distance of one metre, the data therefore indicates that there is a low likelihood of the underwater noise from the backhoe dredging operation causing a disturbance to salmon. The fit to all of the measured data appears low when considering all of the data at each measurement position. It should be borne in mind, however, that the majority of the data is at a lower level, corresponding to periods when the excavator arm is moving to and from the support barge. The higher level fit illustrated in Figure 7-7 has only considered the higher levels of noise at each measurement position, and therefore represents the underwater noise when the excavator arm is extracting material from the seabed. This indicates a worst case Source Level noise for the backhoe dredging operation of 62 db ht (Salmo 1 m, with a propagation loss described by 10 log r. Based on this fit to the data, the noise may remain above 50 db ht (Salmo salar) to a range of approximately 15 m. Beyond this range there is a low likelihood of the underwater noise from the backhoe dredging operation causing a disturbance to salmon. 7.3 Suction dredging Introduction Loosened material is raised to the sea surface in suspension via a pipe system and centrifugal pump. Where the seabed material is loose, the suction of seawater close to the seabed is sufficient to raise material to the sea surface. Firmer material may require mechanical loosening or the use of water jets and hence, suction dredging is most efficient when working with fine substrate materials. A trailing suction hopper dredger is a self propelled ship that is equipped with trailing pipes that suck up seawater and seabed sediment into a large hopper contained within the hull of the vessel. The suction pipe terminates in a drag head that moves over / through the surface of the seabed. At the sea surface, the seawater is allowed to run off, leaving the extracted material in the hopper. Trailing suction dredgers remove layers of material in long runs, and are typically used for channel dredging and harbour maintenance Underwater noise from suction dredging operations. During a break in the backhoe dredging operations at Lerwick, a trailing suction dredger undertaking channel dredging operations passed at a range of 200 m from the survey vessel. Figure 7-8 presents a time history of the underwater noise from this operation. The noise is rather featureless, and is a continuous noise throughout the period of the operation. By listening to the underwater noise recordings it is evident that the dominant components of the noise are from sediments rising up through the suction pipe, characterised by a relatively high frequency broadband hiss. There is also a lower frequency component likely to be ship noise from the dredging vessel. The underwater noise time history illustrated in Figure 7-8 indicates that at a range of 200 m, the noise has a peak to peak level that varies from a minimum of approximately 40 Pa to a maximum of 100 Pa, (152 to 160 db re. 1 µpa). As this is a continuous noise it is more appropriate to specify it in terms of the RMS Sound Pressure. In this case, the analysis of the noise recording indicated that the one second RMS sound pressure varied from 3 to a maximum of 16 Pa over the measurement period, giving a Sound Pressure Level from a minimum of 130 to a maximum of 144 db re. 1 µpa. If spherical spreading of underwater sound is assumed, this would indicate a broadband source level noise for the suction dredging operation of 190 db re. 1 1 m. For comparison, Richardson et al, (1995) quotes a broadband source level of 185 db re. 1 1 m, for this type of operation. 67

74 Figure 7-9 presents the underwater noise from the suction dredging operation in terms of the spectral levels of noise. At a range of 200 m the spectral levels of noise from 40 Hz to 400 Hz exceed 100 db re. 1µPa 2.Hz -1, and remain above 80 db re. 1µPa 2.Hz -1 over the broad frequency range from 25 Hz to 6 khz. Spectral levels of underwater noise at frequencies of 25 khz are typically 30 db above the ambient sea noise, indicating a considerable increase in high frequency noise energy at this range. The low frequency spectral levels are approximately 50 db above the ambient sea noise (at a frequency of 100 Hz, corresponding to frequencies at which fish hear). Analysis of this data in terms of sound perception by the salmon indicated underwater noise at a range of 200 m that varied from 25 to 35 db ht (Salmo salar). Assuming spherical spreading of the underwater noise, this would indicate a source level for the suction dredging operation of approximately 80 db ht (Salmo salar), and a 50 db ht (Salmo salar) loudness zone extending to a range of 50 m. It is concluded that there is a low likelihood of the underwater noise from the trailer suction dredging operations causing a disturbance to salmon. 7.4 Summary of dredging noise impact on salmon Measurements of underwater noise from backhoe dredging operations have indicated that the highest levels of underwater noise occur when the excavator is in contact with the seabed. Based on a 50 db ht (Salmo salar) criteria, during these periods of the backhoe dredging operation, the underwater noise may cause a behavioural disturbance to salmon to a range of 15 m. Analysis of the underwater noise from suction dredging operations has indicated that the 50 db ht (Salmo salar) loudness zone may extend to a range of 50 m. It is concluded that both suction dredging and backhoe dredging activities will cause no significant behavioural disturbance to migratory salmon. 68

75 Range (No. of samples) Sound Pressure Level (db re. 1µPa) RMS db ht (Salmo salar) Max Mean Min Max Min Mean 7 m (n = 279) m (n = 79) m (n = 300) m (n = 230) m (n = 328) m (n = 233) m (n = 100) m (n = 100) Background 2200 m (n = 70) Table 7-1. Summary of the distribution of one second Sound Pressure Levels and db ht (Salmo salar) with range from backhoe dredging operations, Lerwick, 29 th May,

76 Figure 7-1. An underwater noise time history at a range of 7 m from backhoe dredging operations, Lerwick. The two instances of increased underwater noise shown correspond to material being excavated from the seabed. Figure 7-2. An underwater noise time history at a range of 105 m from backhoe dredging operations, Lerwick. The peaks in underwater noise pressure levels during the period when material was being picked up from the seabed are still clearly visible over background noise. Figure 7-3. An underwater noise time history at a range of 2.2 km from backhoe dredging operations, Lerwick. The small variations in underwater pressure levels are due to wave passage. 70

77 Figure 7-4. Variaton in spectral levels of underwater noise with range from backhoe dredging operations, Lerwick. Figure 7-5. Comparison of spectral levels of underwater noise during backhoe dredging operations, Lerwick. 71

78 Figure 7-6. Variation in underwater noise Sound Pressure Level with range from Backhoe dredging operations, Lerwick. (n=1850) Figure 7-7. Variation in measured db ht (Salmo salar) level with range from Backhoe dredging operations, Lerwick. 72

79 Figure 7-8. An underwater noise time history at a range of 200 m from trailing suction dredging operations, Lerwick. Figure 7-9. Spectral levels of underwater noise at a range of 200 m from trailing suction dredging operation, Lerwick. 73