I was recently attending a course on Passive Acoustic Monitoring (PAM) with PAMTech.
They teach people how to use PAMGuard, a vital tool in understanding our marine environment as the software makes analysis of vast amounts of data possible (week/months/years). The software PAMGuard is a community developed modular software tool (managed by SMRU) that I see as complimenting dBSea very well.
While dBSea is focused on the propagation of noise from distinct sources, PAMGuard is focused on the monitoring side and can be used both to measure noise, but first and foremost as a tool to track whales and fish (especially when acoustically tagged).
When doing PAM it is vital to know what volume of water you've actually covered. This can be very complex to estimate for a number of reasons:
So after you have your data PAMGuard can tell you how far away the animal was. However it is not so good as a planning tool, as it will not tell you what your detection range is before you go out to sea.
dBSea can estimate this, and in this way be an important tool for planning surveys.
Depending on what species you want to target on your survey dBSea can help determining your sampling (track) density. In Figure 3 the proposed track is too dense (many redundant measurements) for the low frequency species (baleen whales). The track is probably not bad for the mid-range species (most dolphins), as some redundancy can be good, especially when emitted signals are highly directional (like those used for biosonar). For the highest frequency band, where you'd expect small dolphins and porpoises to be, the track is not dense enough to ensure I have covered the entire volume of water.
(All of the above assumes stationary animals, but in real life they move about faster than the survey boat, wonder how to compensate for that!).
For my scenario in Figure 3, I've assumes source levels of 220 dB re 1μPa. Levels this high are not the norm, but gives you an image of maximal detection range (many whales however will output levels this high! [1-3]). You can also use a detection range plot like the above to estimate total population from the sub-sample your track has registered (if you know a good statistician!).
I would love to see more use of complimenting software to help us get the most out of our effort
Note: I only have some experience with PAMGuard, and I'm sure I only know part of what the programme can do.
Literature:
They teach people how to use PAMGuard, a vital tool in understanding our marine environment as the software makes analysis of vast amounts of data possible (week/months/years). The software PAMGuard is a community developed modular software tool (managed by SMRU) that I see as complimenting dBSea very well.
While dBSea is focused on the propagation of noise from distinct sources, PAMGuard is focused on the monitoring side and can be used both to measure noise, but first and foremost as a tool to track whales and fish (especially when acoustically tagged).
When doing PAM it is vital to know what volume of water you've actually covered. This can be very complex to estimate for a number of reasons:
- Wavelength matters. The low frequency calls from big baleen whales travel very far (>100 km), whistles and mid-frequency-clicks from dolphins not so far (<10 km), and porpoise clicks are so high frequency that attenuation dampens them within 500 meters.
- In shallow water, otherwise far-reaching, low frequency sound can get attenuated greatly because of their tendency to penetrate into the sediment.
- Ducting at the surface or in SOFAR channel can cause sound at some frequencies to travel farther than usually, or cause quiet zones very close to the emitting animal.
PAMGuard is really good at localising animals, from as little as two hydrophones (provided they are towed). It can also utilise the power of many hydrophones (six in Figure 1, ten in Figure 2) to pinpoint animals in 3D.
 |
| Figure 1. Location is given as a probability (colour graded) on rings around a linear array - multiple rings might indicate multiple animals. |
For a better approximation of position non-linear arrays are a must - I only included Figure 1 because I think it looks cooler (more to look at). For a good localisation see below:
 |
| Figure 2. When the array is not linear it becomes possible to localise the animals much more precisely. The insert are the waveforms (clicks) from the ten hydrophones. The location on the left, close to the little blue sphere is the most likely solution (chi-square test). |
dBSea can estimate this, and in this way be an important tool for planning surveys.
Depending on what species you want to target on your survey dBSea can help determining your sampling (track) density. In Figure 3 the proposed track is too dense (many redundant measurements) for the low frequency species (baleen whales). The track is probably not bad for the mid-range species (most dolphins), as some redundancy can be good, especially when emitted signals are highly directional (like those used for biosonar). For the highest frequency band, where you'd expect small dolphins and porpoises to be, the track is not dense enough to ensure I have covered the entire volume of water.
(All of the above assumes stationary animals, but in real life they move about faster than the survey boat, wonder how to compensate for that!).
 |
| Figure 3. An example survey, with receiving ranges in different frequency bands, taking background noise, seabed and water properties into account. Notice that for high frequencies (80-160 kHz) there are "holes" in my survey route. |
I would love to see more use of complimenting software to help us get the most out of our effort
Note: I only have some experience with PAMGuard, and I'm sure I only know part of what the programme can do.
Literature:
- Møhl B, Wahlberg M, Madsen PT, Miller LA & Surlykke A. 2000. Sperm whale clicks: Directionality and source level revisited. Journal of the Acoustical Scosiety of America. 107:638
- Rasmussen MH, Miller LA & Au WWL. 2002. Source level of click from free-ranging white-beaked dolphins (Lagenorhynchus albirostris) recorded in Icelandic waters. Journal of the Acoustical Scosiety of America. 111:1122
- Villadsgaard A, Wahlberg M & Tougaard J. 2007. Echolocation signals of wild harbour porpoises, Phocoena phocoena. Journal of Experimental Biology. 210:56-64
Comments
Post a Comment