Code:46.2008-final-Feb9 Target Strength of Southern Resident Killer Whales (Orcinus orca): Measurement and Modeling Jinshan Xua, Z. Daniel Denga,*, Thomas J. Carlsona, Brian Mooreb a Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352, USA b * BioSonics, Inc., 4027 Leary Way NW, Seattle, WA 98107, USA Corresponding author at: Tel.: +1 509 372-6120; fax: +1 509 372 6089; E-mail address: zhiqun.deng@pnl.gov Abstract— A major criterion for permitting the deployment of tidal turbines in Washington State’s Puget Sound is management of risk of injury to killer whales from collision with moving turbine blades. An active monitoring system is being proposed to detect and track killer whales within proximity of turbines and alert turbine operators to their presence and location to permit temporary turbine shutdown when the risk of collision is high. Knowledge of the target strength of killer whales is critical to the design and application of active acoustic monitoring systems. In 1996 a study of the target strength directivity of a 2.2 m long bottlenose dolphin at an insonifying frequency of 67 kHz was performed. Noting that killer whales, which are dolphins, are morphologically similar to bottlenose dolphins and then assuming allometry, we estimated the relative broadside and tail aspect TS of a 7.5 m long adult killer whale at an insonifying frequency of 67 kHz to be −8dB and −28 dB respectively. We used a three-layer model for plane wave reflection of sound at 200 kHz from the lung of killer whales to estimate their target strength. We assessed the accuracy of our killer whale target strength estimates by comparing them with target strength estimates of free swimming killer whales obtained using a splitbeam active acoustic system operating at 200 kHz. The killer whale target strength estimates based on the preliminary model were in good agreement with those obtained for free swimming killer whales. Keywords - target strength; Southern Resident killer whales; active sonar system; tidal power 1 Code:46.2008-final-Feb9 I. INTRODUCTION Prototype marine hydrokinetic energy devices are in the process of being deployed in coastal waters to evaluate their performance and any environmental consequences of operation. No tidal power generating devices are currently deployed in the United States. A Federal Energy Regulatory Commission preliminary permit has been granted to Snohomish County Public Utility District No. 1 to deploy two tidal turbines built by OpenHydro (Dublin, Ireland) at a site in Admiralty Inlet in Washington’s Puget Sound to study their operation over a 5-year period. A criterion for final approval for deployment of tidal turbines by regulatory authorities is demonstration of capability to manage the risk of injury to killer whales from strike by moving turbine blades. Pacific Northwest National Laboratory researchers are designing and testing a prototype of a marine mammal alert system (MMAS) for detecting the presence of killer whales in the immediate vicinity of tidal turbines. The system will inform tidal turbine operators about the presence of the killer whales to facilitate shut down of the turbines when there is high risk of collision of the whale with the operating turbine. Both passive and active monitoring systems are being considered for the MMAS. The passive monitoring system is being developed by modifying an energybased juvenile salmon acoustic telemetry system (McMichael et al., 2010; Deng et al. 2011; Weiland et al., 2011). Several commercially available sonars are being evaluated for the active monitoring system. An object in water with acoustic impedance different from water scatters a portion of acoustic energy incident from an active acoustic source back toward the insonifying acoustic system or other receivers within range. The ratio of energy incident on the object to that backscattered from the object is defined as the target strength (TS) for the object. With the exception of a sphere of uniform composition, the target strength of an object, such as a fish or whale, is a complex function of the shape of the object and its orientation to the insonifying sonar system. Target strength is a parameter in the sonar equation and is required for the design of a sonar system to optimize design for detection of the targets of interest. The 2 Code:46.2008-final-Feb9 TS of an object is not single valued but a complex function of geometry, size, composition, and frequency of the insonifying sound. The TS of simple geometric shapes with acoustic impedance significantly different from water can be accurately derived analytically. The TS of more complicated shapes with complex composition are typically experimentally determined. However, when measurement of TS is not feasible because of cost or technical constraints, such as the case with killer whales, a combination of analytical analysis and numerical simulation with consideration of the composition of complex objects can be used to estimate TS. Experimental determination of the TS of animals such as fish is typically done in a laboratory where the animal or acoustic measurement system can be manipulated to insonify the whole animal or parts of the animal over a range of aspects (Love, 1971; Foote, 1980; Clay, 1991; Reeder and Stanton, 2004; Reeder et al., 2004). For large animals, such as whales, very few TS measurements have been reported due to the technical difficulty of performing TS measurements on unconstrained animals in the open ocean. The first measurement of a sperm whale was made by Dunn (1969) using an explosive charge as the sound source and a calibrated sonobuoy as the receiver. Dunn reported a TS of −8 dB at 1 kHz for an adult sperm whale. Levenson (1974) reported a similar bistatic maximum TS of 14.4 dB at a center frequency of 12 kHz for sperm whales. Love (1971) measured the TS of humpback whales at a range of about 70 m. Love reported the TS of a 14 m adult humpback insonified at 20 kHz to be 7 dB near side aspect and −4 dB near head side aspect. The TS of a 9 m long juvenile humpback was reported to have a TS of 2 dB near broadside aspect at 10 kHz (Love, 1971). Miller and Potter (2001) reported two TS measurements made at sea for humpback and northern right whales with an 86.25-kHz continuous-wave 3-ms duration pulse transmitted by a phased-array sonar. The broadside TS of an adult humpback whale was observed to be 4 dB. Miller and Potter observed the broadside TS of a 15 m long right whale to be between −4 and −8 dB; they also reported that the head aspect TS of 8 m long juvenile right whales 3 Code:46.2008-final-Feb9 ranges between −7 to −13 dB. Most recently, Lucifredi and Stein (2007) reported an aspect-dependent gray whale TS measurement taken at about 500-m range using a vertical line array sonar frequency centered at 23 kHz. The reported TS measurements for adult gray whales ranged from −2.9 dB at tail aspect to 12.8 dB at broadside aspect. An active sonar system was used to study killer whale hunting behavior in Norwegian waters (Simila, 1997; Nottestad and Axelson, 1999; Nottestad and Simila, 2001; Nottestad et al., 2002a, 2002b). A comprehensive and relatively well-controlled experiment was conducted to insonify a tagged killer whale using two commercial active sonar systems (the Simrad Model SP90 (20-30 kHz) and Simrad Model SH80 (110 to 120 kHz), Kongsberg Maritime AS, Horten, Norway). The goal of the study was to evaluate the feasibility of using active acoustic systems to detect killer whales and observe their behavior in the Lofoten fjords of Northern Norway (Kvadsheim et al., 2007). However, TS observations of killer whales, if obtained during this study, were not reported. Au (1996) measured the TS directivity at 67 kHz of a well-trained bottlenose dolphin in a pool like facility. He also insonified parts of the dolphin to identify the major contributors to whole body TS. Au observed relatively low TS for most portions of the dolphin's body with exception of strong echoes from the lungs. Au’s observations indicate that the air filled lung of whales, and presumably other marine mammals, is a major reflector of sound, which is similar to findings for fish that indicate that the swim bladder is the major reflector of sound in the species of fish that have one (Clay and Horne, 1994). The target strength measurements for large marine mammal at sea are very important but difficult to obtain. Usually, when these measurements are collected, they are most likely only for a specific animal and certain frequency and there are a lot of gaps that we need to fill before we can apply these measurements to practical applications. In this paper, we report the preliminary results of TS modeling we performed to extrapolate the TS measurements of bottlenose dolphin provided by Au (1996) to killer whales. We also report the results of analysis of echo data we acquired for three freely swimming killer 4 Code:46.2008-final-Feb9 whales in 2008 using a 200 kHz split-beam sonar and additional analysis we conducted to explore differences between the in-field TS measurement of killer whales and those we estimated by extrapolation from Au’s data using a 3 layer acoustic scattering model (Miller and Potter 2001). This model assumes that the lung is the major sound reflector in whales and estimates the plane wave backscattering coefficient to account for the acoustic impedance and absorption of whales flesh and blubber at different frequencies. Although it is simple and preliminary, it provides an instructive method for comparing the available measurement of various species at different frequencies and useful information for practical applications in the future. II. A. MATERIALS AND METHODS Measurements In 2008, a DT-X Echosounder (BioSonics, Inc., Seattle, Washington) with a deck unit (BioSonics, Inc.) was deployed to continuously monitor for killer whales at Lime Kiln Point on San Juan Island, Washington State. BioSonics DT-X is a 9˚ (horizontal)× 6˚ (vertical) 200 kHz digital split-beam echosounder, which has been developed and commercially available for the direct measurement of target strength (Carlson and Jackson 1980). The split-beam echosounder has a transducer, which is usually symmetrically divided into four quadrants (Ehrenberg 1974). This BioSonics DT-X echosounder has one large element whose signal is used to determine the echo amplitude, while three smaller elements are used to measure the off-axis angles. The primary goal was to demonstrate the capabilities of autonomous scientific hydroacoustics systems by using an active sonar system to study distribution, abundance, and behavior of killer whales moving past Lime Kiln Point. The first field trial of the monitoring system was carried out from the R/V Gato Verde, a catamaran sailing vessel with better stability than a monohull boat, at a depth of 14 m. The system transducer was 5 Code:46.2008-final-Feb9 mounted on a customized aluminum frame deployed at a depth of 1 m. The coordinates for the location of the transducer were 48° 30.920′ N and 123° 9.168′ W. The transducer was aimed horizontally at a bearing of 238° clockwise from magnetic north. The transducer was connected to the deck unit by power and digital signal cables. The data collected during this field trial are discussed below. The BioSonics DT-X system could be operated in the range from 5 to 250 m at 200 kHz. The system transmitted a rectangular pulse 0.4 ms in duration at a rate of 2 pulses/second. On May 22, 2008, three killer whales (two adults and one juvenile) were visually observed passing by the drifting R/V Gato Verde. The fixed-aim sonar system detected the whales and recorded their approach as shown in the echogram presented in Figure 1. The three rectangular frames (red dashed lines) in the echogram outline three time segments containing sequences of echoes from the visually observed whales. The killer whales were within the beam of the sonar system for about 3 minutes. B. Data Analysis 1) Whale Behavior The killer whale echo sequences were processed to identify the strongest echo returns. The strongest echo returns in each segment that appeared to be from an individual whale were used to estimate the individual whale’s TS. The phase information from the split-beam system for each echo return, when combined with the aiming angles of the system’s transducer, permitted the azimuth and elevation angles to the whales relative to the acoustic axis of the transducer to be estimated. The ranges to the targets were also provided in the system output. While the echoes from individual whales were easily discerned in the echo sequences, the visual observations of the whales were not sufficiently detailed to unambiguously assign a particular whale to 6 Code:46.2008-final-Feb9 each echo sequence or to the echoes selected for estimation of TS. Therefore the target strength estimates discussed below are not specific to particular whales. 2) Target Strength The echo returns from the whales were recorded in the BioSonics .dt4 format (Biosonics, 2008). A quadrature amplitude demodulation technique (Proakis, 2000) was applied to the filtered analog echo signals out of the system transducer to provide a digital record of the echo returns with a sampling frequency of 41.667 kHz. The target strength was estimated by compensating the beam pattern in the relevant direction relative to the on-axis direction. The detection threshold was set at −70 dB re 1μPa to allow the target to be detected only in the direction on the beam axis. 3) Aspect Direction/Incident Angle The split-beam system processed the echo returns to estimate the azimuth and elevation of received echoes relative to the transducer’s aiming direction. The angle of incidence of the transmitted pulse on a whale was estimated by computing the angle of travel of the whale relative to the acoustic axis of the system transducer from changes in range to the whale on successive echo returns. When range was increasing the whale was assumed to be near tail aspect and when range was decreasing the whale was assumed to be near head aspect. Similar methods have been used by others, specifically Levenson (1974) and Lucifredi (2007). 4) Target Strength Modeling of Killer Whales at 67 kHz Assuming allometry in lung size between species of dolphins we used the TS directivity data of Au (1996) for a bottlenose dolphin to estimate the aspect dependent TS for killer whales. In Au’s study, a female bottlenose dolphin weighing 126 kg and measuring 2.2 m in length was trained to remain stationary at a ―bite‖ station while it was insonified to determine its TS. The transducers of the 7 Code:46.2008-final-Feb9 measurement system were manipulated to obtain the TS directivity (i.e., aspect dependent TS) of the dolphin. Three different signals were used to measure the animal’s TS (two frequency-modulated pulses and a broadband click with a peak frequency of 67 kHz). The target strength estimates at each aspect were averages from echoes of 20 pings acquired for each angle of insonification of the whole dolphin or of specific portions of the dolphin (Au, 1996). The ratio of the lung length to the full body length of this dolphin was about 0.286. Assuming that dolphins and killer whales share a similar morphological structure and that lung size is proportional, a 7.5 m long killer whale (mature female) would have a lung length of about 2.143 m. Given the difference in TS of a 0.6-m-diameter sphere and 2.143-m-diameter sphere is about 10.77 dB (Urick, 1983) we estimate that a 7.5 m long killer whale will have a TS at 67 kHz about 10.77 dB higher than that for a 2.2 m long bottlenose dolphin. 5) Three-Layer Model To Extrapolate Target Strength Modeling of Killer Whales at 200 kHz Miller and Potter (2001) suggested the difference in TS estimates between gray and right whales was due to the difference in the thickness of their blubber layers. They also suggested an approach to estimate TS at other frequencies. Their method requires estimates of blubber thickness, its density and the speed and attenuation of sound through the blubber. The mathematic formula for three-layer reflection coefficients can be found in Medwin and Clay (1997), Miller and Potter (2001), and Brekhovskikh and Lysanov (2003). The TS of killer whales at higher frequencies such as 200 kHz can be estimated based on the plane wave reflection coefficient as a function of frequency for the elements of a three-layer model (skin, blubber layer, and lung) for a whale (Miller and Potter, 2001). Miller and Potter (2001) estimated the reflection coefficient for different thickness of blubber using a compressional sound speed of 1700 m/s, density of blubber of 1200 kg/m3, and attenuation of sound through blubber of 1 neper/m. In contrast, Jaffe et al. (2007) measured the attenuation of 100 kHz sound through sirenia (manatee) blubber to be about 5.57 dB/(10MHzmm), 8 Code:46.2008-final-Feb9 which equals 6.64 neper/m. For our analysis we chose to use values of of 4.45 neper/m at 67 kHz, 1.53 neper/m at 23 kHz, and 13.28 neper/m at 200 kHz. The plane wave reflection coefficients for a three-layer model of water–blubber–lung at three different frequencies are shown in Figure 2 for an angle of incidence of 90 normal to the surface of the blubber. It also shows the variability in the reflection coefficient as a function of blubber thickness, which increases with frequency. Depending on the thickness of blubber layer, the difference in reflection coefficient between the frequencies of 67 kHz and 200 kHz could vary from 4 dB up to 15 dB. These reflection coefficient values can be used to estimate the TS of killer whales at 200 kHz. III. A. RESULTS Whale Behavior Figure 3 shows the averaged distance, depth, and swimming speed estimates for the killer whales observed a Lime Kiln Point. There were some unrealistic values for whale depth (above surface), possibly due uncertainty in the tilt angle of the split beam transducer in the horizontal direction or propagation effect such as multipath off the ocean surface of the transmitter pulse and/or echo. Other than these observations, the trends in whale depth appear to be realistic given the visual observations of the whales. The whales were initially detected by the sonar at a range of about 100 m, swimming at a depth of about 2 or 3 m. They approached to within about 30 m from the sonar transducer at a depth of about 5 m. At this range from the transducer they turned and swam away at a deeper depth. The killer whales were also observed to slow on their approach to the transducer then speed up after they turned and were swimming away from the transducer. 9 Code:46.2008-final-Feb9 B. Target Strength Temperature and salinity were also measured by the BioSonics DT-X system and used to improve the accuracy of sound speed estimation. The resulting estimated speed of sound used in the echo ranging calculations was 1475.68 m/s. Killer whale TS was estimated using the echo sounder source level and the two-way transmission loss for 200 kHz sound in sea water. To improve the TS estimation, the relative angles between the sending acoustic beam axis and returning echo beam axis were used to compensate for the narrow beam effect by computing the composite beam pattern for the transmitting beam and the angle of arrival of the echo. During the 3-min the whales were observed by the sonar system (Figure 3), the whales initially approached the transducer at a fairly high swimming speed for about 20 s and then slowed and approached closer to the transducer over a 45 s period. In the final 20 s of observation, the whales turned from the transducer and swam away at higher speed until no longer detected by the sonar system. In each period (as annotated in Figure 1 with red color frames), the incident angles should have different statistics, because the aspect of insonification of the whales would have differed. During approach the whales would have been insonified at near head aspect; during their turn at a range of aspects from near head to broadside and finally near tail, and when swimming away at near tail aspect. For this reason, we analyzed the data in these three sections separately. Figures 4 illustrates the TS as a function as distance and time and provides histograms of TS data for each. In the first section, the averaged TS was −33 dB and the averaged speed was about 2.1 m/s. In the second section, the averaged TS was about −31 dB with an averaged speed of 1.6 m/s; it had highest TS measure of −4 dB. In the third section, the averaged TS was −23 dB with an averaged speed of 2.7 m/s. The three sections of observations where the whales were behaving differently and the aspect of the whale to the sonar system differed had different TS statistics. The first one looked like a central-weighted 10 Code:46.2008-final-Feb9 Gaussian distribution. The second one had a long tail at higher TS values. The TS in the third section has a bimodal distribution. The differences in the TS statistics for each segment are most likely due to the differences in the aspects at which the whales were insonified and the complexity of swimming whales as acoustic targets. The first segment had more TS measured at head aspect because the whales were approaching the transducer. In the second segment the whales presented more of a broadside aspect to the sonar system initially followed by more tail aspect. In the third segment the whales were at more of a tail aspect. In the next section we will consider in more detail the observed whale TS as a function of aspect to the insonifying acoustic beam. C. Target Strength As Function of Aspect Direction/Incident Angle Figure 5 shows histograms of the estimated TS and angle of incidence of the insonifying beams on the whales for the three observation segments. In each set of histograms the frequency of occurrence of observed TS values are in the upper panel and histograms for the frequency of occurrence of estimated insonifying aspects are given in the lower panels. In the first set of histograms (Figure 5a-b) that correspond to the first observation segment where the whales were approaching the sonar transducer, most of the incident angles (lower panel) are less than 100° (0 is head aspect, 90 is broadside aspect, 180 is tail aspect); there are few observations above 100°. In the second set of histograms corresponding to the second observation segment (Figure 5c-d), the lower panel shows the incident angles spread across the entire range from head to tail aspect but with higher frequency of occurrence at broadside aspect (90). The third set of histograms for the third observation segment when the whales were moving away from the sonar transducer (Figure 5e-f), the incident angles are more frequent from broadside toward tail aspect. These estimates of the aspect of 11 Code:46.2008-final-Feb9 insonifying beam incidence are consistent with our interpretation of whale motion based on analysis of the change in range with time of echoes from the whales. D. Target Strength As Function of Aspect Angle The TS observations of three individual of killer whales as a function of the angle of incidence of the acoustic system beam are shown in the polar plots of Figure 6 for the three observation segments. In the first observation segment, most of the TS measurements are for head-on aspect because the whales were approaching the transducer. The TS observations for this segment vary from about −45 dB at an insonifying incident angle of 20° (near head) to −22 dB at an aspect angle of 80° (near broadside aspect) (Figure 6a). In the second observation segment (Figure 6b), the highest TS observations, about −10 dB, were at 90° incidence (broadside); however, there are also a few observations of high TS at 0° (head aspect) and 180° (tail aspect). In the third observation segment (Figure 6c), the TS measurements were primarily from incident angles larger than 90°, more toward tail aspect, as expected for whales that were moving away from the sonar system. The highest TS value observed was −7 dB at broadside aspect. The data also show a lobe in the directivity at about 110°, with TS values up to −13 dB. The lowest TS values observed were near −40 dB. E. Target Strength Modeling of Killer Whales The TS directivity estimates at 67 kHz for killer whales along with the TS directivity for bottlenose dolphins measured by Au (1996) are shown in Figure 7. The estimated TS of killer whales is modeled on the measurements of bottlenose dolphin reported by Au (1996) with a 10.77 dB offset to account for the difference in the size of the lungs in the two dolphin species. The TS of an adult killer whale 7.5 m in length at a frequency of 67 kHz is approximately −8 dB at the broadside aspect and −28 dB from the tailon aspect. This estimate is applicable for 7.5-m killer whales swimming close to the surface; however, the TS of aquatic mammals vary with water depth because of changes in lung volume with increasing 12 Code:46.2008-final-Feb9 pressure. The change in TS with depth could be estimated assuming that lung volume would decrease by 50% for each doubling of hydrostatic pressure (Ridgeway and Howard, 1979). IV. DISCUSSION AND CONCLUSIONS Field measurement of the TS of marine mammals in general and large whales in particular is very difficult for many reasons. Data from the May 2008 deployment of a BioSonics 200 kHz active sonar system at Lime Kiln Point provided a unique opportunity to estimate the TS of killer whales. The analyses of these data provided estimates of killer whale TS over insonifying aspects from tail to head. The TS measurements obtained at 200 kHz of killer whale TS can be compared with TS estimates obtained using a preliminary three-layer model of acoustic reflectivity at 67 kHz. The Southern Resident killer whale preys on adult Chinook salmon in Puget Sound and elsewhere in the Pacific Ocean near the northwestern Pacific coast. Killer whales should develop a blubber layer but not as thick as that of right whales or even humpbacks. In terms of overall size, killer whales are relatively smaller than right whales (10.7 to 16.8 m long) and humpbacks (12.2 to 15.2 m long). Thus, assuming killer whale blubber thickness is about 0.1 m, approximately −10 dB must be added to the TS model of killer whales TS at 67 kHz. This adjustment to the 67 kHz estimate of TS directivity for killer whales would result in 200 kHz TS values ranging between −18 dB to −38 dB depending upon aspect angle. The TS estimates obtained so far are for a 7.5 m killer whale. However, the adult male killer whales can be as much as 10 m in length. Given our proportionality arguments we would expect the TS of larger killer whales to be as much as 3 dB higher than those for the smaller 7.5 m long whale. This would increase the range of TS for the larger killer whales to −15 dB to −35 dB. The measurement data for the free swimming killer whales at Lime Kiln from the first observation segment range from −22 dB to −45 dB where the whales were believed to be observed at near head on aspect. Measurement data from 13 Code:46.2008-final-Feb9 the second observation segment yielded TS values ranging from −10 dB to −48 dB over a range of aspect from head through broadside to tail. In observation segment three, TS observations ranged from −7 dB to −40 dB for retreating whales insonified at the tail aspect. These TS measurements are relatively comparable to the model for killer whales at 200 kHz, which predicts TS values within the range from −15 dB to −35 dB. To our knowledge, no other field measurements of killer whales TS have been reported in the literature. The target strength model of killer whale TS is based on TS measurement of a bottlenose dolphin at 67 kHz and a three-layer reflection model used to extrapolate modeled killer whales TS values from 67 kHz to 200 kHz. The TS of an adult killer whale was modeled for an insonifying frequency of 67 kHz by assuming that a killer whale shares similar morphological structure with a bottlenose dolphin and that the ratio of lung length to total length is the same across dolphin species and sizes of animals within species. The TS of killer whales at higher frequencies - 200 kHz were modeled using a simplified three-layer model with correction based on the estimated differences in acoustic reflection coefficients as a function of blubber thickness at different frequencies. Interestingly, the averaged measurement data of three different Lime Kiln observation segments show that the TS of killer whales are in the range of −11 dB (broadside) to −43 dB (tail aspect). The modeling result gives range from −18 dB (broadside) to −38 dB (tail aspect) at 200 kHz frequency. Both the modeling and measurements TS results are surprisingly low. In Au’s 1996 measurement of the bottlenose dolphin, he found an unexpected decrease in TS at the broadside aspect as the frequency increased from 23 to 45 kHz. Au explained the TS decrease might be due to the anechoic properties of dolphin skin or blubber layer because, based on the formula provided in Urick (1983), the TS for most underwater objects increases or remains constant with frequency. However, the data used by Urick (1983) are mostly for simple geometric shapes of uniform composition and are assumed to have a large impedance difference from water. None of these assumptions hold for the 14 Code:46.2008-final-Feb9 body of whales, which are complex, composed of layers with very different physical and acoustic properties. Even when the three-layer model is simplified by not considering the skin and muscle, it might still provide a better estimate of acoustic reflectance than the two-layer model with large impedance mismatch used by Urick. These considerations might explain the unexpectedly low TS measurements Au’s 1996 obtained for bottlenose dolphin and might also help explain the relatively low killer whale at 200-kHz frequency. As with any other experiment, unavoidable measurement errors are always a factor affecting estimates. For example, the pulse length of the Biosonics DT-X system is 400 μs (user selectable from 0.1 ms to 1.0 ms), which is likely too short to insonify an entire whale; this is most certainly the case when the whale is approaching or swimming away from the sonar. In that case, the TS of killer whale at tail or head aspects would be underestimated. The TS of an underwater object is likely to be less at short ranges than at long ranges. One reason is that a very directional active sonar signal fails to insonify the entire target. Another reason is the target reflection as echo usually acts differently from point source, depending on the size and shape of the underwater object. The trajectory of whale motion in Figure 1 and the whales’ speed as a function of time in Figure 3 show that during the 3-min sonar observation period, the whales first were approaching the transducer for about 20 s and then moved closest to the transducer for about 45 s. In the final 20 s, the whales moved away with increased speed. Similar behavior was observed also in studies by both Love (1971) and Lucifredi and Stein (2007) with center frequencies of 20 kHz and 23 kHz for humpback whales and gray whales. They observed that the whales approached the transmit site, slowed down or even stopped to listen for a while, and then moved on with increased speed. It is generally considered safe to use echosounders at frequencies above 180 kHz in the vicinity of marine mammals because this is out of the hearing range of most marine mammals. However, the recent study of a signal leakage of active sonar 15 Code:46.2008-final-Feb9 systems (Deng et al., 2010) shows that a signal other than the primary signal at lower frequency could potentially fall into the hearing range of marine mammals (Southall et al., 2007), which is consistent with the observation in this case that the killer whales were capable of hearing the 200 kHz sonar because of transmit signal spectral components at frequencies within the hearing range of killer whales. Acknowledgment The work described in this article was funded by The Wind and Water Power Program of the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy. The study was conducted at Pacific Northwest National Laboratory (PNNL) in Richland, Washington, which is operated by Battelle for the U.S. Department of Energy. The authors thank Tim Acker of BioSonics Inc. for providing the original data. We also thank Charlie Brandt, Andrea Copping, Jayson Martinez, Graysen Squeochs, and Andrea Currie of PNNL, who provided comments and technical help in preparing the manuscript. 16 Code:46.2008-final-Feb9 References Au, W.W.L. 1996. Acoustic reflectivity of a dolphin. J. Acoust. Soc. Am. 39:3844–3848. BioSonics. 2008. BS&E-2008-04-0009-1.4- DT4 Data File Format Specification. BioSonics, Inc., Seattle, Washington. Brekhovskikh, L. M., and Lysanov, Yu. P. 2003. Fundamentals of Ocean Acoustics, 3rd ed. Berlin: Springer. 278 pp. Carlson, T.J. and Jackson, D.R. 1980. Empirical evaluation of the feasability of split beam methods for direct in situ target strength measurement of single fish. Seattle Applied Physics Laboratory, University of Washington, 43 pp. Clay, C. S. 1991. Low-resolution acoustic scattering models: fluid-filled cylinders and fish with swimbladders. J. Acoust. Soc. Am. 89:2168–2179. Clay, C. S., and Horne, J. K. 1994. Acoustic models of fish: the Atlantic cod (Gadus morhua). J. Acoust. Soc. Am. vol. 96(3):1661–1668. Deng, Z. D., Weiland, M. A., Seim, T. A., LaMarche, B. L., Choi, E. Y., Fu, T., Carlson, T. J., and Eppard, M. B. 2011. A cabled acoustic telemetry system for detecting and tracking juvenile salmon: part 2. three-dimensional tracking and passage outcomes. Sensors 11(6):5661–5676. 17 Code:46.2008-final-Feb9 Deng, Z., M. A. Weiland, Fu, T., Seim, T. A., LaMarche, B. L., Choi, E. Y., Carlson, T. J., and Eppard, M. B. 2010. Design and instrumentation of a measurement and calibration system for an acoustic telemetry system. Sensors 10(4):3090–3099. Dunn, J. L. 1969. Airborne measurements of the acoustic characteristics of a sperm whale. J. Acoust. Soc. Am. 46: 1052–1054. Endangered Species Act of 1973. 7 U.S.C. Section 136, 16 U.S.C. Section 1531, et seq. Ehrenberg, J.E. 1974. Two applications for a dual-beam transducer in hydroacoustic fish assessment systems. Proceedings of the IEEE International Conference on Engineering in the Ocean Environment, 21–23 August 1974, IEEE 1974, pp. 152–5. Foote, K. G. 1980. Rather-high-frequency sound scattering by swimbladdered fish. J. Acoust. Soc. Am. 78:688–700. Jaffe, J. S., Simonet, F., Roberts, P.L.D., and Bowles, A. E. 2007. Measurement of the acoustic reflectivity of sirenia (Florida manatees) at 171 kHz. J. Acoust. Soc. Am. 121:158–165. Kvadsheim, P., Benders, F., Miller, P., Doksæter, L., Knudsen, F., Tyack, P., Nordlund, N., Lam, F.-P., Samarra, F., Kleivane, L., and Godø, O. R. 2007. Herring (sild), Killer Whales (spekkhogger) and Sonar – the 3S-2006 Cruise Report with Preliminary Results. FFI-rapport 2007/01189, Forsvarets forskningsinstitutt/Norwegian Defence Research Establishment (FFI), Horten, Norway. 18 Code:46.2008-final-Feb9 Levenson, C. 1974. Source level and bistatic target strength of the sperm whale (Physeter catadon) measured from an oceanographic aircraft. J. Acoust. Soc. Am. 55:1100–1103. Love, R. H. 1971. Dorsal-aspect target strength of an individual fish. J. Acoust. Soc. Am. (49):816– 823. Lucifredi, I., and Stein, P. J. 2007. Gray whale target strength measurements and the analysis of the backscattered response. J. Acoust. Soc. Am. 121:1383–1391. Marine Mammal Protection Act of 1972. 16 U.S.C. Sections 1361-1421h, October 21, 1972, as amended 1973, 1976-1978, 1980-1982, 1984, 1986, 1988, 1990, 1992-1994 and 1996. McMichael, G. A., Eppard, M. B., Carlson, T. J., Carter, J. A., Ebberts, B. D., Brown, R. S., Weiland, M., Ploskey, G. R., Harnish, R. A., and Deng, Z. D. 2010. The Juvenile Salmon Acoustic Telemetry System: a new tool. Fisheries 35(1):9–22. Medwin, H., and Clay, C. S. 1997. Fundamentals of Acoustical Oceanography. New York: Academic Press. 712 pp. Miller, J. H., and Potter, D. C. 2001. Active high frequency phased-array sonar for whale ship strike avoidance: target strength measurements. In: MTS/IEEE OCEANS 2001 – An Ocean Odyssey, vol. 4, pp. 2104–2107. Washington, D.C.: Marine Technology Society. 19 Code:46.2008-final-Feb9 Nøttestad, .L, and Axelsen, B. E. 1999. Herring schooling manoeuvres in response to killer whale attacks. Can. J. Zool. 77:1540–1546. Nøttestad, L., and Similä, T. 2001. Killer whales attacking schooling fish: why force herring from deep water to the surface? Mar. Mamm. Sci. 17:343–352. Nøttestad, L., Ferno, A., and Axelson, B. E. 2002b. Digging in the deep: killer whales’ advanced hunting tactic. Polar Biol. 25:939–941. Nøttestad. L., Ferno, A., Mackinson, S., Pitcher, T. J., and Misund, O. A. 2002a. How whales influence herring school dynamics in a cold front area of the Norwegian Sea. ICES J. Mar. Sci. 59:393–400. Proakis, J. G. 2000. Digital Communications, 4th ed. New York: McGraw-Hill. 1024 pp. Reeder, D. B., and Stanton, T. K. 2004. Acoustic scattering by axisymmetric finite-length bodies: an extension of a two-dimensional conformal mapping method. J. Acoust. Soc. Am. 116:729–746. Reeder, D. B., Jech, J. M., and Stanton, T. K. 2004. Broadband acoustic backscatter and highresolution morphology of fish: measurement and modeling. J. Acoust. Soc. Am. 116:747–761. Ridgeway, S. H., and Howard, R. 1979. Dolphin lung collapse and intramuscular circulation during free dive. Science 206:1182–1183. 20 Code:46.2008-final-Feb9 Similä, T. 1997. Sonar observations of killer whales (Orcinus orca) feeding on herring schools. Aquat. Mamm. 23(3):119–126. Southall, B. L., Bowles, A. E., Ellison, W. T., Finneran, J. J., Gentry, R. L., Greene Jr., C. R., Kastak, D., Ketten, D. R., Miller, J. H., Nachtigall, P. E., Richardson, W. J., Thomas, J. A., and Tyack, P. L. 2007. Marine mammal noise exposure criteria: initial scientific recommendations. Aquat. Mamm. 33(4):411–522. Szymanski, M. D., Bain, D. E., Kiehl, K., Pennington, S., Wong, S., and Henry, K. R. 1999. Killer whale (Orcinus orca) hearing: auditory brainstem response and behavioral audiograms. J. Acoust. Soc. Am. 106:1134–1141. Urick, R. J. 1983. Principles of Underwater Sound for Engineers. New York: McGraw-Hill. 423 pp. Weiland, M. A., Deng, Z. D., Seim, T. A., LaMarche, B. L., Choi, E. Y., Fu, T., Carlson, T. J., Thronas, A. I., and Eppard, M. B. 2011. A cabled acoustic telemetry system for detecting and tracking juvenile salmon: part 1. engineering design and instrumentation. Sensors 11(6):5645–5660. 21 Code:46.2008-final-Feb9 List of Figures FIGURE 1 Echogram acquired with a BioSonics split-beam DT-X system on May 22, 2008 at Lime Kiln State Park that shows three time periods containing echoes from killer whales. 22 Code:46.2008-final-Feb9 FIGURE 2 Plane wave reflection coefficient as a function of blubber thickness for a three-layer model of water– blubber–lung at three frequencies: 23-kHz, 67-kHz, and 200-kHz. 23 Code:46.2008-final-Feb9 FIGURE 3 The averaged distance from the sonar system transducer (a), swimming depth (b), and swimming speed (c) of three killer whales. 24 Code:46.2008-final-Feb9 FIGURE 4 Target strengh estimates for three killer whales. The panels in the left column (a: Section 1; c: Section 2; e: Section 3) show target strength observations for a killer whale as a function of distance from the sonar system transducer and time. The panels in the right column (b: Section 1;d: Section 2; f: Section 3) show a histogram of the whale target strength estimates. 25 Code:46.2008-final-Feb9 26 Code:46.2008-final-Feb9 FIGURE 5 Histogram of the target strength (upper pannels) and acoustic beam incident angles (lower row panels) for the three observational segments when whales were observed approaching (5a-b), turning away from (5c-d), and retreating from the acoustic system transducer (5e-f). 27 Code:46.2008-final-Feb9 FIGURE 6 Target strength of three killer whales as function of acoustic beam incident aspect angles. (a) Observation segment 1; (b) Obervation segment 2 ; (c) Observation segment 3. 28 Code:46.2008-final-Feb9 FIGURE 7 Polar plot of the estimated target strength of killer whale at frequency of 67 kHz. The red curve is the target strength of bottle-nose dolphin measured by Au (1996). 29