The PAMGUARD project was set up to provide a world standard software infrastructure for acoustic detection, localisation and classification of marine mammal vocalisations to enhance acoustic monitoring as part of mitigation procedures and for research into marine mammal abundance, distribution and behaviour.
Many marine activities involve underwater sound emissions. These may be produced as a by-product of the activity (e.g. piling or explosives), or intentionally for sensing (e.g. air guns used for seismic surveys in oil and gas exploration, or military/commercial sonar). Marine mammals are particularly sensitive to sound underwater and this leads to concerns that they might be physically affected or their hearing might be affected if they are exposed to high levels of sound. One strategy for mitigating against these risks is to search for animals within a zone of influence and either delay or shut down noise producing operations if sensitive animals are detected within this zone. This is sometimes called operational mitigation.
Why Use Acoustics?
The default method for detecting marine mammals at sea is to look for them. Visual observations play a vital role, but marine mammals are difficult to spot on the sea surface, especially when weather and light conditions are poor. In addition of choose, visual techniques are next to impossible at night but often operators wish to continue noise producing activities round the clock. Many marine mammals produce loud and distinctive vocalisations, however and these preset alternative and often complimentary method of detection. Often, marine mammals vocalise most when they are away from the surface and not available to be seen. In addition, acoustic cues can often be detected more reliably at greater ranges and are less affected by weather and sighting conditions and animals can be detected acoustically equally well day and night. Passive Acoustic Monitoring isn't a panacea but for many species it can significantly increase the probability that they are detected and increase the effectiveness of mitigation.
Why Automate Detection and Localisation?
Many marine mammal vocalisations can be heard perfectly well by human monitors so why go to so much trouble to provide software to automate the process? There are several good reasons.
1. The first is that marine mammal voclise over a huge range of frequencies. Much greater than the human auditory range. Thus, within the cetaceans, blue whales and other balaenopterids produce infrasonic vocalisations below the 20Hz lower bound of human sensitivity while small odontocetes are high frequency specialists. Harbour porpoise for example produce narrow band pulses in the high ultrasonic at around 130kHz. Thus, equipped only with our ears we would pick up only a narrow subset of available vocalisation.
Click for more information on marine mammal acoustics
2. A second reason is that even if when they aren't quite as good as the human ear computer detection algorithms are consistent, they don't get distracted, bored or tired and they provide a reliable and repeatable level of monitoring
3 A third reason is that computer assistance to provide a degree of automation can reduce the size of team required. This can be particular important in offshore monitoring scenarios where berth space is limited and the cost of providing observers is high.
4. Finally, advanced signal processing linked with sophisticated hydrophone arrays can provide data such as 3D locations that a simple human observer listening on headphones could never achieve.
Why did we need PAMGUARD?
Good acoustic monitoring software existed before PAMGUARD but there were a number of reasons that justified developing something new.
In the first place, it was realised that there was a real value in having a single software that marine mammal observers (MMOs) could become familiar with and use on a variety of different vessels. Ideally that software should be freely available, interface to a wide range of hardware configurations and work on many different computer platforms. (Pamguard achieved cross platform compatibility by being written in Java.)
None of the existing programs were open source. This meant that the functioning and performance of the algorithms within them was often not clear and it wasn't possible for a group of users to contribute to and to support it. There was also a long term risk that the software might be withdrawn from use or become outdated.
In most cases there was no commitment to supporting and updating the software and as it wasn't open source it would be difficult for other programmers to provide such support.
Some of the software, though excellent, was not designed for real time monitoring by a single operator in field conditions.
To help users of existing software, and to avoid reinventing wheels, much of the functionality, as well as the look and feel of some of the more popular existing real time monitoring software, the IFAW Logger suite and Ishmael.
PAMGUARD Conference and Workshop
A conference and series of discussion workshops dealing with many aspects of passive acoustic monitoring and the PAMGUARD project was held at Heriot-Watt University, Edinburgh, in March 2007.