2
Sources of Sound in the Ocean and Long-Term Trends in Ocean Noise

INTRODUCTION

In this chapter the major natural (physical and biological) and anthropogenic contributors to ocean noise are discussed. Gaps in our knowledge or available data are identified that will need to be addressed in future research in order to develop predictive models of the effects of noise on marine mammals. A more thorough description of modeling efforts is contained in Chapter 4.

This chapter focuses on the properties of the sources and does not describe in detail the effects on the environment as the acoustic energy travels away from the vicinity of the sources. Parameters such as source level (in units of dB re 1 µPa at 1 m), source spectral density level (units of dB re 1 µPa2 per Hz at 1 m), and time-integrated source pressure amplitude squared for use with transient signals (units of dB re 1 µPa2 at 1 m) are presented for many of these sources, particularly man-made sources. However, accurate estimation of the source properties for many types of naturally occurring sounds is impossible, given the lack of knowledge of the individual source locations, of the spatial distribution of multiple contributing sources, and of the complex propagation conditions. Therefore, in such situations, the measured properties of the received acoustic field (which are obtained directly and require no additional information, computation, or assumptions, but which contain the effects of propagation) will be presented. The text clearly differentiates between the properties of the sources and those of the received field. The distinction between source level and received level also is discussed both in Chapter 1 and in the Glossary.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 27
2 Sources of Sound in the Ocean and Long-Term Trends in Ocean Noise INTRODUCTION In this chapter the major natural (physical and biological) and anthropogenic contributors to ocean noise are discussed. Gaps in our knowledge or available data are identified that will need to be addressed in future research in order to develop predictive models of the effects of noise on marine mammals. A more thorough description of modeling efforts is contained in Chapter 4. This chapter focuses on the properties of the sources and does not describe in detail the effects on the environment as the acoustic energy travels away from the vicinity of the sources. Parameters such as source level (in units of dB re 1 µPa at 1 m), source spectral density level (units of dB re 1 µPa2 per Hz at 1 m), and time-integrated source pressure amplitude squared for use with transient signals (units of dB re 1 µPa2 at 1 m) are presented for many of these sources, particularly man-made sources. However, accurate estimation of the source properties for many types of naturally occurring sounds is impossible, given the lack of knowledge of the individual source locations, of the spatial distribution of multiple contributing sources, and of the complex propagation conditions. Therefore, in such situations, the measured properties of the received acoustic field (which are obtained directly and require no additional information, computation, or assumptions, but which contain the effects of propagation) will be presented. The text clearly differentiates between the properties of the sources and those of the received field. The distinction between source level and received level also is discussed both in Chapter 1 and in the Glossary.

OCR for page 27
In the absence of shipping, natural forces are the dominant sources of the long-term time-averaged ocean noise at all frequencies. In the presence of distant shipping, contributions from natural sources continue to dominate time-averaged ocean noise spectra below 5 Hz and from a few hundred hertz to 200 kHz. The dominant source of naturally occurring noise across the frequencies from 1 Hz to 100 kHz is associated with ocean surface waves generated by the wind acting on the sea surface. Nonlinear interactions between ocean surface waves called microseisms (see the Glossary; referred to as “Surface Waves—Second-Order Pressure Effects” in Plates 1 and 2) are the dominant contributors below 5 Hz, while thermal noise (i.e., the pressure fluctuations associated with the thermal agitation of the ocean medium itself) is the dominant contributor above 100 kHz. Natural biological sound sources make a noticeable contribution at certain times of year. For example, a peak around 20 Hz created by calls of large baleen whales is often present in deep-ocean noise spectra. Groups of whistling and echolocating dolphins can raise the local noise level at the frequencies of their signals. Snapping shrimp are an important component of natural noise from a few kilohertz to above 100 kHz close to reefs and in rocky bottom regions in warm shallow waters. Fish can add to ocean noise in some locales. Whether intentional or unintentional, anthropogenic noise in the marine environment is an important component of ocean noise. Sound is a widely used tool for a broad range of marine activities. In the search for new hydrocarbon reserves, the rock underlying the seafloor is characterized using air-guns. Marine researchers use sound waves to investigate the properties of seawater both for local and global studies. Sonars used for civilian navigation and defense purposes use sound waves to locate objects under the sea surface. Unintentional contributions to marine noise arise from transiting ships, coastal and marine construction activity, mineral extraction, and aircraft overflights. These anthropogenic sound sources contribute to ocean noise over the complete 1-Hz to 200-kHz band of interest in this report. In the lowest bands, 1-10 Hz, the contributors are ship propellers, explosives, seismic sources, and aircraft sonic booms. In the 10-100 Hz band, shipping, explosives, seismic surveying sources, aircraft sonic booms, construction and industrial activities, and naval surveillance sonars are the major contributors. For the 100-1,000 Hz band, all the sources noted for the 10-100 Hz band still contribute. Also, the noise from nearby ships and seismic air-guns can extend up into the 1,000-10,000 Hz band. This band also includes underwater communication, naval tactical sonars, seafloor profilers, and depth sounders. The 10,000-100,000 Hz band includes the systems listed, in addition to mine-hunting sonars, fish finders, and some oceanographic systems (e.g., acoustic Doppler current profilers). Anthropogenic contributors at and above 100,000 Hz are limited to mine hunting, fish finders, high-resolution seafloor mapping devices

OCR for page 27
such as side-scan sonars, some depth sounders, some oceanographic sonars, and research sonars for small-scale oceanic features (Table 2-1a and 2-1b). Prior to considering anthropogenic sources, it is useful to first understand the natural sources that contribute to ocean noise. Presumably, hearing and communication systems of marine organisms are adapted to these natural noises. NATURAL SOURCES OF OCEAN NOISE Physical and Geophysical Sources The ocean is intimately coupled to the solid earth and the atmosphere, and in fact, most of the significant physical sources of natural sound occur at the interfaces among these three media. Additional sound in the marine environment originates in the atmosphere and penetrates the ocean surface. Elastic vibrations in the earth also introduce sound into the underwater acoustic field. Sources at the Ocean Surface The dominant physical mechanisms of naturally occurring sound in the ocean occur at or near the ocean surface. Most are associated with wind fields acting on the surface and the resulting surface wave activity. In the absence of man-made, biological, and transient sounds, ambient noise is wind dependent over the band from below 1 Hz to at least 50 kHz. Below 5-10 Hz, the dominant ambient noise source is the nonlinear interaction of oppositely propagating ocean surface waves. These sounds are called microseisms. (The term “microseisms” comes from the fact that they also are the dominant source of noise in high-quality, on-land seismometer measurements; however, the source mechanism for microseisms is unrelated to seismic processes in the solid earth.) Across most of the remainder of this band, the primary sources are bubbles that are oscillating, both individually and collectively in a cloud, in the water column. Several good references on natural physical sources of ocean noise and the properties of the ambient noise field are available (e.g., Urick, 1984; Zakarauskas, 1986; Ross, 1976; Kerman, 1988, 1993; Buckingham and Potter, 1995; Leighton, 1997; Deane, 1999). Only a brief summary of the major contributors to the underwater sound field is given here. However, in some frequency bands such as the band from 10 to 200 Hz, where ambient noise in the northern hemisphere typically is dominated by shipping noise, the dominant source mechanisms have not been identified. Quantification of the relative contributions of the various mechanisms of naturally occurring sound created at the sea surface remains an active area of research. The average ocean noise spectrum can be empirically described and

OCR for page 27
TABLE 2-1a Characteristics of Anthropogenic Contributors to Marine Noise SOURCE SPATIAL VARIABILITY DIRECTIONALITY ACTIVITY LEVELS   Large Scale Ocean Basin to Global Mid Scale 10s of km to Ocean Basin Small Scale <1 km to 10s of km   Number of Regional Sources Frequency of Activity in Region Shipping Presence is global for all types and limited to the ocean surface   Merchant Shipping lanes transcend ocean basins and are populated continuously   1-2 4-5/hr Utility   Operations are confined to subocean basins and localized areas such as fishing grounds   1-30 daily       All shipping: Generally considerd to be omnidirectional, but shielding is certainly present in the horizontal plane, especially for the higher frequencies; omnidirectional in the vertical plane.   Military   Operations are military exercises, war zones Extending down to amphibious assault zones, beach heads   6-10 bi-monthly

OCR for page 27
Scientific   Specific sites to observe phenomena of limited spatial scales Down to localized phenomena such as “black smokers”   1-2 monthly Recreation     Coastal regions, limited range   >10 >5/day Other E.g., transoceanic cable laying Localized operations E.g., drill site     >monthly Seismic exploration   Surveys to >100 km Down to 10s of km Omnidirectional 1 monthly Sonars Global presence, but variability is defined by sonar use   Military   Surveillance Ocean basin use Down to 10s of km   Omnidirectional 1 monthly Tactical   10s of km and up, conditions permitting   Horizontal plane, Vertical >100°, Vertical plane <20° 2-3 See host platform data above Weapon/ Counter Weapon   10 m to >10 km Highly directional in both planes (<5°) 1-2   Civilian   Communications   10s of km and up Down to >1 km Horizontal plane: omni Vertical plane: <10° 1  

OCR for page 27
SOURCE SPATIAL VARIABILITY   DIRECTIONALITY ACTIVITY LEVELS   Large Scale Ocean Basin to Global Mid Scale 10s of km to Ocean Basin Small Scale <1 km to 10s of km   Number of Regional Sources Frequency of Activity in Region Navigation   <1 km to <10 km Omnidirectional 1   Hi-resolution   >10 m to >100 m Highly directional (almost all look down or up) 1-2   Marine Research Limited ocean basin tests Spatial interest to ocean basin dimensions Down to sub-meter measurements See sonars / military and civilian 1 monthly Explosions   <1 km Omnidirectional 1 seldom Industrial Activity Presence is global and limited to near-shore locations/onshore locations   Construction   <1 km to 10s of km Omnidirectional 1   Dredging <1 km to 10s of km Omnidirectional 1 Power plants 100s of m Omnidirectional 1 Factories 100s of m Omnidirectional 1

OCR for page 27
Transportation   10s of km upward <1 km to 10s of km Road noise – unknown, Ferries – same as ships 1–2 Up to 1/hr Miscellaneous Global presence   Aircraft overflight Very similar in extent and character to shipping lanes   See text-confined to vertical cone 1 4-5 hrs Military activity nonsonar   < 1 km Omnidirectional 1 seldom Note: Spatial variability is an indicator of the geographic distribution of the sources. Directionality refers to the directi on in which the signal is projected. Activity levels indicate the likely number of regional sources, and the frequency of signal occurrence wi thin that region.

OCR for page 27
TABLE 2-1b Characteristics of Anthropogenic Contributors to Marine Noise SOURCE TEMPORAL VARIABILITY   SOURCE CHARACTERISTICS     Large Scale wks to mos Mid Scale hrs to days Small Scale sec(s) to min(s) Signal Structure Spectral Content Source level dB re 1µPa at 1 m Shipping   See Figure 2-1 for example   Merchant Constant presence   occasional transient due to operations activity on vessel   160–220 Utility On site for wks Down to hrs   numerous transients due to nature of operations For all shipping, broadband energy from 10 Hz to >1 kHz with spectral lines rising above B/B due to propulsion blades, turbines, generators, etc. 160−200   Transient additions are broadband and flat from 10 Hz to 1 kHz 160−200

OCR for page 27
Military   On site for hrs to days   general level up and down with exercise/war fighting requirements   160–220 Scientific On site for days Down to min(s) stop and start behavior driven by data collection schedule 160–200 Recreation Other Wks Typically hrs days /hrs Or less highly variable 160–190 Seismic exploration   On site for days   Impulsive (see Figure 2-4) Broadband >240 Sonars   Military   Surveillance On site for wks Down to days   Pulsed tones <1 kHz >230 Tactical   On site for hrs Down to min(s) Pulsed tones >1 kHz to <10 kHz 200 to 230 Weapon/ Counter Weapon hrs to ~ a day Down to min(s) Pulsed tones / Wideband pulses >0 kHz to >100 kHz 190 to 220

OCR for page 27
SOURCE TEMPORAL VARIABILITY   SOURCE CHARACTERISTICS     Large Scale wks to mos Mid Scale hrs to days Small Scale sec(s) to min(s) Signal Structure Spectral Content Source Level dB re 1 µPa at 1 m Civilian   Communications   hrs To min(s) tones CW/Pulsed Low kHz to >10 kHz 180-210 Navigation   hrs/days min(s) CW/Pulsed Low kHz to >10 kHz 180-210 Hi-resolution   min(s) Pulsed tones >10 kHz to >100 kHz 160-220 Acoustic   Series of pulses 10-500 msec w/ interpulse periods of 0-10 sec 5-30 kHz typically 130-150 Harassment or Deterrent Devices (AHD, ADD)   Marine Research   Down to days/hrs   See military/civilian sonars See military/ civilian sonars 160–220 Explosions   sec(s) Impulsive (see figure) Broadband >240

OCR for page 27
Industrial Activity   Construction wks to mos Down to days   Broadband and Tones/CW <10 Hz to 1 kHz Unknown Dredging wks to mos Down to days Broadband and Tones <10 Hz to <1 kHz Unknown Power plants Constant process   CW and some transients <100 to several 100 Hz Unknown Factories Constant process CW and significant transients <100 to several 100 Hz Unknown Transportation Constant process Same as shipping and highway noise See shipping and broadband 170–210 Miscellaneous   Aircraft overflight Constant process   Both CW and broadband <100 Hz to 10 kHz Unknown Military activity nonsonar   hrs Down to sec(s) Impulsive and broadband Broadband Unknown Note: Frequently reported source signal characteristics are given although additional source characteristics, such as rise time, may also be important in determining the effects of the sources on marine mammals. This table is not meant to be a catalog nor does it approach all-inclusiveness but is provided to give a sense of the breadth of human impact on the undersea environment.

OCR for page 27
FIGURE 2-6 The spectrum of the acoustic signal from an underwater explosion. The quantity labeled “energy flux density” actually is the instantaneous pressure amplitude squared summed over the duration of the signal, as discussed in the Glossary. To convert the units on the right-hand vertical axis from dB re 1 [(dynes/ cm2)2](sec) at 1 yd into dB re 1 (µPa2)(sec) at 1 m, approximately 100 should be added to the values so that the resulting axis extends from 162 to 222 dB re 1 (µPa2)(sec) at 1 m. These spectral levels pertain to a 1-lb. charge detonated at a depth of 20 fathoms (36.6 m) and are equal to the actual source level at each frequency for a signal of 1-sec duration. The corresponding broadband zero-peak pressure level at 1 m from the source for the initial shock wave from a 1-lb. charge of TNT is 272 dB re 1 µPa at 1 m, as given by Equation 2-1 in the text. The plot shows the addition of the shock-wave and bubble pulse energies at frequencies greater than 1/T, with T equal to the time interval between the shock wave and the first bubble pulse. SOURCE: Urick, 1975. Reprinted with permission from the Acoustical Society of America. SL(0–pk)(dB re 1 µPa at 1 m) = 271.8 dB + 7.533*log(w) (2-1) where w is the charge weight in pounds. The third column of Table 2-5 provides some examples from the use of Equation 2-1. Industrial and Construction Contributions to Marine Noise The range of activities in this category is extremely broad, ranging from power plants located near the seaside to pile driving, dredging, shipyards, canal lock structure operations, and general harbor daily functions. The coupling of this energy, which is a combination of terrestrially based to shoreline and nearshore, into the marine environment is poorly understood.

OCR for page 27
TABLE 2-5 Zero-to-Peak Pressure Level and Spectral Level at 1 kHz of Pressure Amp Squared Times Duration for High Explosive Detonated at 40 m Depth TNT (lb.) Spectral Level at 1 kHz of Pressure Amp Squared Time Duration (dB re 1 µPa at 1 m) Zero-to-Peak Pressure Level at 1 m (dB re 1 µPa2s at 1 m) 1 192 272 10 200 279 100 207 287   SOURCE: Urick, 1975. Courtesy of McGraw-Hill. This broad range of activities produces a range of source levels and acoustic patterns: pile driving (impulsive, very high amplitudes), power plants (very strong 60-cycle and harmonics), industrial (tones at frequencies of machinery operating speeds), dredging [both shipborne machinery and mechanical motion (suction and earth-moving devices, possible explosive use)], and power-generating windmills. A typical spectral structure is broadband with the superposition of a number of lines originating from reciprocating machinery or engines. Some measurements of the underwater sounds created by these types of sources have been presented in the open literature (e.g., Richardson et al., 1995). Additional measurements are contained in various technical reports and memoranda. It would be useful to gather these measurements together into one easily accessible place (as the committee recommends) for use by the scientific and regulatory communities as well as others. Note, however, that the coupling of land-based vibrations and very nearshore sounds into the offshore underwater acoustic field is highly dependent on the geology, morphology, and length of the land-based portion of the propagation path. Therefore, measurements made in one offshore area are not necessarily applicable to other offshore areas. Numerical modeling of the coupling between land-based vibrations and the ocean acoustic field is a subject of current research. This uncertainty associated with the coupling process makes an assessment of the overall impact of these sounds on the marine environment difficult. However, of greater importance is to understand the potential impact of a given noise source in its actual geological setting on the marine ecosystems that are located nearby. At present, the evaluation for land-based and very nearshore sources probably is best done using actual underwater acoustic measurements in the region of interest.

OCR for page 27
LONG-TERM TRENDS IN OCEAN NOISE One of the most important and challenging issues that emerges in an examination of ocean noise and its effects on marine life is the quest to determine any long-term trends in the overall levels of sound in the sea. How has noise in the sea increased with time since the 1850s through increased industrialization and related maritime activities? What parameters, other than direct noise measurements, might be related to the overall sound levels produced by the myriad of sources described? What, if any, modeling capabilities exist to predict ocean noise levels and other noise characteristics in the decades to come? Is there any hope that humans might influence these predicted changes through time by introducing appropriate mitigation measures? Answering these questions holds enormous significance for life in the sea; after all, long-term changes in background noise levels may influence animal behavior and impact the very existence of a particular species. Long-term trends are particularly insidious in that they result from the gradual accumulation of effects over much shorter periods of time for which these effects may appear to be imperceptible. Although the importance of assessing long-term trends in ocean noise levels is clear, there is a remarkable dearth of theories or data addressing this topic. Commercial shipping noise is apparently the only area in which it is possible to make informed comments concerning long-term trends, and even in that case, the data sets are very limited and the discussion is usually speculative. The focus on shipping implies an emphasis on frequencies of a few hundred hertz and below and a geographical bias toward the northern hemisphere, where most of the dominant shipping lanes exist. In this section, the first attempts to estimate the preindustrial noise background by examining measurements in areas of the South Pacific with extraordinarily low ship traffic are described (Cato, 1997, 2001). The addition of the anthropogenic component of noise during the Industrial Revolution, principally the result of shipping, is reviewed, followed by a summary of existing data on long-term trends in shipping noise levels and a discussion of various indicators for evaluating and predicting shipping noise levels. Finally, speculations on the long-term trends in ocean sounds other than those from shipping are presented. Recommendations for future research to measure and predict long-term trends in ocean noise are listed in Chapter 5. Preindustrial Noise One approach to modeling long-term trends is to hypothesize that the overall background noise level remained essentially constant until the onset of the Industrial Revolution in 1850. At that time, land-based industrial activities began to escalate rapidly and resulted in an enormous increase in the use of ships under power to transport goods and provide services over

OCR for page 27
the oceans. Other related, though less significant, developments were the powering of the world’s naval vessels and the expansion of coastal and offshore construction activities. Additional, though also secondary, sources of anthropogenic noise that emerged much later (primarily during the past 50 years) were those produced by offshore oil exploration and drilling activities, naval sonars, and acoustical oceanographic research. This model is based on the assumption that the noise contributed by natural physically generated and biological sources is independent of time and that human-generated noise prior to the Industrial Revolution was negligible. In fact, this assumption is open to debate, since there is some evidence that global climate change effects have resulted in higher sea states (Bacon and Carter, 1993; Graham and Diaz, 2001); these could potentially cause an increase in the noise levels generated by breaking waves over time. This effect is very likely of secondary importance, however, and therefore the model in which commercial shipping provides the primary time-dependent influence on long-term noise levels is adopted here. The waters surrounding Australia provide a unique opportunity to estimate preindustrial noise levels. Shipping densities in some areas are extremely low or negligible. This situation provided Cato (1997b, 2001) with the opportunity to determine the “usual lowest ocean noise” level from an extensive suite of measurements (Plate 5). There are several striking features that appear in these data. First, the lowest noise level decreases monotonically from 55 dB re 1 µPa2 per Hz at 10 Hz to 30 dB re 1 µPa2/Hz at 11 kHz. Second, the ship traffic noise data indicate values as high as 80 dB re 1 µPa2/Hz at 20 Hz in the Tasman Sea, “and these approach the traffic noise levels of North American and European waters” (Cato, 2001), consistent with the Wenz curves (cf. Plate 1). In fact, at 20 Hz the ship noise level exceeds Cato’s lowest level by about 25 dB. On the other hand, in the Timor and Arafura seas, the ship noise levels vary from 50 to 58 dB re 1 µPa2 per Hz in the band 20-200 Hz and exceed the lowest level by only a few decibels. Third, Cato (2001) points out that the level of naturally generated noise, both physical and biological, frequently equals or exceeds that produced by ship traffic at 200 Hz and below. Specifically, the Australian data indicate that the wind wave noise continues to increase below 200 Hz, in contrast to the behavior of the deep-water Wenz curves5 for various sea states, which are based on northern hemisphere data. Cato suggests that this is probably due to the difficulty of separating the effects of ships and breaking waves in northern waters. 5   Shallow-water measurements, including those of Wenz (1962), Piggott (1964), and Arase and Arase (1967), indicate an increase in noise levels produced by wind waves at frequencies below 100 Hz.

OCR for page 27
Postindustrial Noise With the Cato data providing a glimpse at the preindustrial noise background, it is now appropriate to examine any existing data on long-term trends in noise levels. As mentioned earlier, the paucity of data on this topic is surprising. However, it is encouraging that the few existing data sets are consistent with one another. First, consider the data and interpretation provided by Ross (1976, 1993). Long-term trends in ambient noise levels can be observed at low frequencies (unspecified, but presumably below 200 Hz) in the East Pacific and East and West Atlantic Oceans (Figure 2-7). Ross (1993) concluded from these data that “low-frequency noise levels increased by more than 10 dB in many parts of the world between 1950 and 1975,” corresponding to about 0.55 dB per year. This increase was attributed to two factors associated with commercial shipping, namely a doubling of the number of ships, which accounts for an increase of 3-5 dB, and greater average ship speed, propulsion power, and propeller tip speed, which are responsible for at least an additional 6 dB. Ross (1976) also speculated that, during the next quarter century, ship noise levels may increase by only about 5 dB because “the number of ships may be expected to increase only about 50 percent and the noise per ship by only a few dB.” FIGURE 2-7 Long-term trend for low-frequency ambient levels for period 1958– 1975. SOURCE: Ross, 1993, courtesy of Acoustics Bulletin.

OCR for page 27
FIGURE 2-8 Point Sur autospectra compared with Wenz (1969). Point Sur data are converted to one-third octave levels and then normalized by the third-octave bandwidths for direct comparison. Shown for reference are the “heavy” and “moderate” shipping average deep-water curves presented by Urick. SOURCE: Andrew et al., 2002. Reprinted with permission of the Acoustical Society of America. In another attempt to assess long-term trends, Andrew et al. (2002) compared noise measurements made on a receiver on the continental slope of Point Sur, California, from 1994 to 2001 with those collected on the same receiver from 1963 to 1965 (Wenz, 1969). The results of their analysis indicated an increase of approximately 10 dB over 33 years (about 0.3 dB per year) from 20 to 80 Hz (Figure 2-8). Andrew et al. attributed this change principally to increases in the number and gross tonnage of commercial ships, a conclusion consistent with Ross’s results. They indicated that they do not have a satisfactory explanation for the increased noise from 100 to 400 Hz (up to 9 dB) or the minimum increase of 3 dB close to 100 Hz. Mazzuca (2001) synthesized the results of Ross (1976) and Wenz (1969) to obtain an overall 16-dB increase in low-frequency noise level from shipping during the period 1950-2000. This value corresponds to a rate of increase of 0.32 dB per year, or about 3 dB per decade.

OCR for page 27
Indicators for Evaluating and Predicting Shipping Noise How does one quantitatively correlate ocean noise levels with shipping activity and its origins in industrial activities? Efforts to determine the principal sources of noise on ships have constituted an active area of research for quite some time. The classical model, put forward by Ross (1976), states that ship-radiated noise is directly related to ship length and speed. Yet this theory was criticized recently by Wales and Heitmeyer (2002), who contend that there is a “negligible correlation between the source level and the ship speed and the source level and the ship length.” Part of the reason for this lack in correlation may be due to the type of source model used by Wales and Heitmeyer (Heitmeyer, personal communication, 2002). In any case, these observations complicate attempts to determine the principal ship parameters affecting the overall noise levels associated with a large number of ships. Nevertheless, it is possible to make some general well-founded comments regarding ship-radiated noise and shipping traffic and their possible implications for long-term ocean noise levels. First, there is no doubt that ships generate noise, principally by propeller cavitation and machinery. Second, it is well known that aging ships tend to generate more noise as mechanical and electrical systems deteriorate over time. Third, newer ships have a number of noise-mitigating characteristics, including quieter diesel-electric propulsion systems and deeper propellers that are less prone to cavitation. Fourth, and most important, the number of ships and gross tonnage of the world fleet have increased substantially since 1950 (Figure 2-9) (McCarthy and Miller, 2002). During this period, the number of ships almost tripled (from 30,000 to 87,000 ships), while the gross tonnage increased by a factor of about 6.5 (from 85 to 550 million gross tons). Interestingly, the logarithmic (dB) equivalent of a factor of 6.5 is about 16 dB, exactly corresponding to the observed increase in low-frequency noise levels. These data suggest the following simple relationship between changes in noise levels and gross tonnage: (2-2) While Equation 2-2 is highly speculative and its predictive capability must be tested (other parameters, such as changes in ship speed, may need to be included, although they may all be correlated with gross tonnage), there is no doubt that world economic conditions influence shipping activity, which in turn affects overall noise levels in the sea. Westwood et al. (2002) estimate “that over 90 percent of world trade is carried by sea and over the period 1985 to 1999, world seaborne trade increased by 50 percent to

OCR for page 27
FIGURE 2-9 Global shipping fleet trends, 1914–1998. Only those commercial ships registered in the U.S. (the U.S. flag fleet) are subject to U.S. regulations and laws when operating outside U.S. territorial waters. SOURCE: McCarthy, 2001. Courtesy of http://coultoncompany.com. about 5 billion tons with the largest increase coming in crude oil and oil products shipments. During 1990-1998 growth averaged 3.2 percent per annum.” The 50 percent trade increase is comparable to the 38 percent increase in gross tonnage during the same period. Applying Equation 2-2 gives a result, 20*log(1.38) = 2.8, which is not much different from the

OCR for page 27
expected increase in noise levels of 4.5 dB (equal to 0.32 dB/year multiplied by 14 years) for this same time period. Only further study can elucidate whether the similarity in these figures is purely coincidental or scientifically meaningful. Long-Term Trends in Other Sources of Ocean Noise No long-term systematic ocean acoustics data set exists to permit a scientific assessment of trends of noise in the ocean. Therefore, the following discussion speculates on possible trends rather than describing any. Although the levels of naturally occurring sound from physical sources (particularly wind-generated and ice-generated noise) may be changing as a result of possible changes in weather patterns associated with global warming, these changes are believed to be dwarfed by other trends. Ocean measurements do exist that demonstrate that ambient sound from some biological sources is increasing in a few locations in the world, for example, sounds produced by humpback whales in the waters around Australia (Cato and McCauley, 2002). The noise associated with whales is expected to asymptote at preexploitation levels as the whale populations return to their preexploitation numbers. However, the overall trend in noise from all biological sources is unknown. Regarding anthropogenic noise sources, the previous sections of this chapter show that educated speculation (Ross, 1976; Mazzuca, 2001) and measurements at one location (Andrew et al., 2002) suggest that shipping noise at low frequencies (20-80 Hz) has increased by about 10-15 dB over a 25-50-year period. Although the decrease in detection range associated with this increase in noise can be calculated from a sonar systems perspective, the degree to which this change has an adverse impact on the marine environment is unknown. The change in level itself is not a cause of great concern given that naturally occurring processes can change noise levels by 20-30 dB over short periods (e.g., Plate 1). However, other properties of this increase in shipping noise may be biologically important, such as the increase in the prevalence of noise (decrease in time intervals between shipping-noise-dominated periods or increase in the number of locations where shipping noise is a significant contributor), the character of the shipping-generated signals themselves, and so on. Increases in the number and size of commercial and recreational craft have resulted in noise-level increases substantially greater than 10 dB in some areas (e.g., 30 dB or so in the frequency band from 10 to 100 Hz in Singapore Harbor) (Potter and Delory, 1998), but the potential impact on these environments is unknown. This trend has been accompanied by the proliferation of boats and ships equipped with depth sounders and fish finders, which have likely raised the high-frequency (above a few kilohertz) noise in some localized areas. Again, the amount of increase and its potential effects are unknown.

OCR for page 27
Trends in seismic exploration are much simpler to define in terms of activity than in terms of contribution to the underwater sound field. As discussed previously in this chapter, industry publications periodically report the numbers of surveys presently being conducted in general locations. However, given that exploration methods have been changing, for example, large explosive sources have been replaced by air-guns, which have evolved into air-gun arrays that focus the radiated acoustic energy in the vertical direction, and that undiscovered oil and gas reserves probably are deeper within the earth and/or are to be found in deeper waters, the overall impact of changes on the ocean sound field is difficult to evaluate without a combined ocean noise measurement and numerical modeling effort. In contrast to focusing of acoustic energy in the vertical by present-day geophysical exploration sources, the newly developing low-frequency navy sonars radiate acoustic energy preferentially in the horizontal direction. Because of the very low absorption of sound in the ocean at low frequencies, these active sonar signals can travel over large distances. Another recent trend in U.S. military sonar has been toward the use of active systems in coastal and shallow-water regions. These sonars have the potential to adversely impact marine mammals; evidence indicates that navy mid-frequency (1-10 kHz) tactical sonars were directly related to the March 2000 mass stranding of marine mammals in the Bahamas (Evans and England, 2001). Other mass strandings have been associated with the transmission of high-level sonar signals, for instance, the May 1996 event in the Mediterranean Sea (D’Amico and Verboom, 1998). Whatever the frequency band, the growth of sonar activity for military purposes started from essentially zero to the present-day levels over just the last half century or so. The number and type of man-made explosions that affect the world’s oceans also have been changing. The largest of these events is associated with nuclear tests, which have taken place only since 1945 (Lawson, 2002). Many of the early tests by the United States from 1945 to 1962 were atmospheric tests conducted on small islands in the central part of the Pacific Ocean just north of the equator. The energy released during these tests certainly created high-level impulsive ocean acoustic signals that traveled over great distances. Much of the subsequent U.S. testing was done underground, sufficiently far from the coast that little impact on the ocean environment occurred. However, underground tests near coastlines can create high levels of underwater sound. As an illustration, one of the French nuclear tests in 1995-1996 on the Mururoa Atoll in French Polynesia in the South Pacific Ocean generated underwater signals that were recorded by a single omnidirectional hydrophone at a range of 6,670 km with levels 20-45 dB above background noise across the frequency band 2-30 Hz (D’Spain et al., 2001). Given the steady progress since the mid-1990s in the number of nations that have signed, and have ratified, the Comprehensive Nuclear Test Ban Treaty, these tests appear to be increasingly rare.

OCR for page 27
Long-term trends in the use of chemical explosive devices also may be taking place. Any speculation on these trends must exclude times of war (the underwater noise created by explosions during the great naval battles in World War II must have been extremely high) since the occurrence of war and resulting contributions to the ocean noise field are highly unpredictable and extremely episodic. Long-term trends in the use of smaller explosive devices also may be taking place. As mentioned previously, explosive sources used in seismic exploration are being replaced by air-gun arrays. However, explosives are routinely used to sever abandoned well-heads so that they can be removed and to decommission the rigs themselves. As oil production in a given region matures and declines, the use of explosives in this way increases. The use of explosives in ocean acoustic and geophysical research has decreased, but these sources still are deployed in a few experiments. Military use of explosive charges as the source component in active sonar systems (e.g., SUS; Urick, 1975) appears to be decreasing. Hull shock tests are rare events and do not appear to be changing significantly in frequency of occurrence. The use of seal bombs has been discouraged by U.S. and international regulations and is being replaced by other types of acoustic deterrent devices. Fishing by the detonation of underwater explosives (a technique whose success improves with its increasing adverse impact on the marine environment) is banned but still is practiced in some regions. In any case, one quantitative measure of the long-term change in numbers and spatial distribution of underwater explosions is possible to obtain, at least for the North Pacific Ocean. The number and estimated source locations of detonations recorded over a modern-day period of time could be compared to those recorded by 20 Missile Impact Location System hydrophones over a one-year period from August 1965 to July 1966 (following Spiess et al., 1968). Nearly 20,000 explosions were detected within this one-year period, with the winter rate of occurrence of 300 explosions per month increasing to 4,000 explosions per month in summer. The highest activity was detected off the west coast of North America, in the Gulf of Alaska, north of Hawaii, and seaward of the Japanese and Kuril Islands (Spiess et al., 1968). The significance of any of these possible changes in the occurrence of underwater explosions to the marine environment is unknown.