Oceans14 Taipei Harumi etal

1 downloads 98 Views 547KB Size Report
Nov 19, 2013 - Narora barrage (downstream) in the up-river district of the. Ganges ..... [8] H. Sugimatsu, J. Kojima, T. Ura, R. Bahl, S.K. Behera and V.S. Sagar,.
Development of a handy mobile 4-hydrophone array system for in-situ census of the Ganges river dolphins (platanista gangetica) Harumi Sugimatsu1, Junichi Kojima2, Tamaki Ura3, Katsunori Mizuno1, Akira Asada1, Rajendar Bahl4, Sandeep Behera5, Hari Singh5, Vivek Sheel Sagar5,Rupak De6 (1) Institute of Industrial Science, The University of Tokyo, Tokyo, Japan, [email protected] (2) KDDI R&D Laboratories Inc., Fujimino, Japan (3) Kyushu Institute of Technology, Kitakyushu, Japan (4) Indian Institute of Technology Delhi, New Delhi, India (5) WWF-India, New Delhi, India (6) PCCF and Chief Wild life Warden of Uttar Pradesh Abstract—Around 15-20 Ganges river dolphins (platanista gangetica) inhabit the 12-km stretch from Karnavas to Narora Barrage in the Ganges river system. However, more than 10 dolphins were carried away downstream after the flood of JuneJuly 2013. In order to urgently conduct a quick census to secure the dolphins in the original habitat, the existing compact fourhydrophone array system (J-Array) was remodeled for towing, and a function to display the real-time acoustic data on a GUI (Graphical User Interface) on an operational PC at the site was added. Using the new J-Array system, the initial in-situ census in combination with acoustic and visual census was conducted for two days from 19 to 20 November 2013. The functions that were added to the new J-Array performed well. Immediately following the towing survey, using both the acoustic and visual census data, an estimate of the number of the dolphins was made at the site. It was estimated that a minimum of six dolphins were still in the habitat. Off-line acoustic data analysis results were also found in accordance with the in-situ census results. From this, the effectiveness of the new J-Array system for conducting in-situ census has been demonstrated. Keywords-Ganges river dolphin, Passive Acoustic Monitoring (PAM), in-situ census, acoustic and visual census.

I.

INTRODUCTION

The Ganges river dolphin (Platanista gangetica) belongs to the small toothed whale (dolphin) family and inhabits mainly the Ganges river system in India. Similar to other dolphin species, the Ganges river dolphin uses bio-sonar clicks and the returning echoes to understand its surrounding environment and to capture prey [1], [2]. The dolphins are threatened owing to ongoing human activities. For the observations conducted for the conservation of the Ganges river dolphins, therefore, field studies in the wild are required. However, since they inhabit mostly turbid water and stay underwater except for respiration, it makes visual observations very difficult. Passive acoustic monitoring (PAM) using arrays of hydrophones that can record the high frequency echolocation clicks of the dolphins has been used as one of the major observation systems for field studies in the wild environment [3], [4]. Because of no adverse effects on the dolphins and

their habitat, a PAM system using one or more arrays of hydrophones provides position information of the click sound sources with their acoustic parameters such as waveform, frequency, ASL (Apparent Source Level), and ICI (Inter-click Interval). These parameters are helpful for understanding the underwater behavior of the Ganges river dolphins and their biosonar click characteristics. The authors have been developing various types of hydrophone array systems that can record the bio-sonar clicks of the dolphins and estimate their 3-D position and have conducted field observations using the developed systems since 2006 [5], [6]. Based on these experiences, targeting 15-20 Ganges river dolphins (adults, young adults and calves) that inhabit a 12-km river stretch from Karnavas (upstream) to Narora barrage (downstream) in the up-river district of the Ganges river system, long-term in-situ monitoring has been ongoing at the stationary observatory that consists of multiple array systems since 2008 (Note: more than 10 dolphins were carried away downstream after the flood of June-July 2013) as shown in Fig. 1 [7], [8]. Though the dolphins migrate several kilometers up and down the river stretch almost everyday, the observable detection range of the stationary observatory is limited by the baseline of LBL (Long Base Line) system used by the arrays. For understanding the trends of the diurnal migration behavior of the dolphins along the 12-km river stretch, a periodic census covering the entire habitat along the river is necessary. In addition, after the natural disasters such as floods or storms, a rapid census should be quickly conducted to secure the dolphins in the habitat. Line transects are widely used for the visual census of wild animals. However, detection misses for the dolphins are unavoidable. In addition, sighting probability of the surfacing dolphins for respiration depends on the ability of the observers and the lighting and weather conditions. Therefore, an effective combination of acoustic and visual census is becoming more common for dolphin census [9], [10]. However, acoustic census using the existing equipments such as a data logger system requires analyzing processing time.

To increase the dolphin’s detection probability, use of a combination of acoustic and visual census, and introduction of a real-time acoustic data display function at the site is efficient because it can achieve in-situ correlation between the acoustic census data and visual census data. Consequently, based on the existing small J-Array system developed for the stationary observatory in 2008 [11], the authors have developed a handy mobile 4-hydrophone array system (new J-Array) in 2012 that consists of a four-hydrophone array, a signal processing unit, and a battery unit as shown in Fig. 2. A GPS (Global Positioning System) receiver and wireless LAN (Local Area Network) are also installed in the system for displaying the dolphin’s real-time location data on a GUI (Graphical User Interface) on a PC in time domain and also in L/L (Latitude and Longitude). After operational tests of the new J-Array over short distances along the river stretches of the Ganges river system and the Mahakam river system (Borneo) in December 2012 and 2013, respectively, we targeted some Ganges river dolphins that stayed on in the 12-km river stretch from Karnavas to Narora Barrage after the above river flood. The initial census conducted in combination with the visual survey was conducted for two days from 19 to 20 November 2013, wherein the 6th phase long-term in-situ monitoring at the stationary observatory was also started (c.f. Fig. 1). In, this paper, the system development for the new J-array and operational results of the initial in-situ census results obtained in combination with both the acoustic and visual census are reported.

Figure 1. Arrangement of the multiple array systems

II.

A. Subject Around 15-20 Ganges river dolphins (adults, young adults and calves) inhabit the 12-km stretch from Karnavas to Narora Barrage in Ganges river system. However, more than 10 dolphins were carried away downstream after the flood of June-July 2013. Based on the visual observations after the flood, WWF India estimated that 6-7 dolphins (an adult (female), young adults, calves over one year old) stayed on in the 12-km river stretch from Karnavas to Narora Barrage. To secure the dolphins in the habitat after the flood, a census for obtaining an accurate number of the dolphins that still remained upstream was to be made urgently. For this purpose, an in-situ census method using the new J-Array along with visual observation for the above Ganges river dolphins has been studied. B. New J-Array system Fig. 2 shows a photograph of the new J-Array. It consists of a four-hydrophone array, and a signal processing unit that are deployed from the side of a support boat, along with a battery unit that is on the support boat during the operation. The length of the four-hydrophone array and the signal processing unit is 0.685 meters and the diameter of the hydrophone array unit is 0.20 meters. The length of the battery unit is 0.24 meters and its diameter is 0.09 meters. Two underwater connectors for antennas of GPS and wireless LAN are attached on the end cap of the cylinder of the signal processing unit. Fig. 3 shows the system architecture of the new J-array, and Table I shows the specifications of the new J-Array. The dolphin click signals received by the hydrophones are amplified, filtered, digitized by 16 bit 500kHz A/D converters, and transferred into the DSP (Digital Signal Processor) memory using DMA (direct memory access). After judging the existence of a click by the signal level and its spectrum components, time differences between the hydrophone signals are calculated using cross-correlation. The acoustic direction of the sound source is calculated from these time differences and is stored in the SD flash memory with additional information such as time-stamp, signal levels, and spectrum components. The waveform also can be recorded. In addition, a GPS (Global Positioning System) receiver and wireless LAN (Local Area Network) were also installed in the system for displaying the dolphin’s real-time direction data on a GUI (Graphical User Interface) on a PC in time domain and also in L/L (Latitude and Longitude). (a)

Figure 2. New J-Array structure

MATERIALS AND METHODS

(b)

(5) The above equations can be used when -90