CAROL STEPIEN: Now we have Joe giving us a focus on hydrophones, characterization of ocean sound to inform the conservation and management of US Marine ecosystems. Joe Haxel is from our Newport lab. 

JOE HAXEL: All right. Thanks, Carol. Yeah, I'm here to tell you guys about the keystone project PMEL was running using a unique hydrophone network to inform the conservation management of US marine ecosystems by characterizing ocean sound levels across US basin. So, ocean sound is an important component for healthy marine ecosystems. Many marine mammals, fish, crustaceans rely on acoustics for a variety of life functions would it be communication, navigation, foraging, protection of territory. 

Over the last several decades, low-frequency sound has been shown to be increasing in parts of the Northeast Pacific, and this has been attributed to an increase in commercial shipping throughout the basins. The impact from anthropogenic noise disturbances aren't very well understood. So these are masking our loss of communications space between negative behavioral responses like [INAUDIBLE] habitat, and in extreme cases, there can be a negative physiological response to [INAUDIBLE]. 

To give you guys a feeling for what anthropogenically dominated soundscape might feel-- sound like versus a natural sound. I have a couple of recordings from our network. So first is from a naturally dominant sound [INAUDIBLE] on a fringing reef in Western Pacific in Samoa. You're going to hear some snapping from humpback whales vocalizing. [INAUDIBLE] And the next one is an anthropogenically dominated soundscape. And if you're sensitive to sounds, this is going to be pretty loud. But this is emphasizing the point I'm trying to make. [INAUDIBLE]

So that was also a humpback whale near the shipping lanes off the Boston Harbor. So in response to this rising concern, National Fisheries Service rolled out an agency-wide ocean noise strategy. The PMEL Acoustics Group has served as an implementation arm of science goals of that strategy and providing the hydrophone technology, the mooring technology, and the operation of the network is also including some data analysis. 

So we address a couple of key questions. Number one, trying to fill those information gaps by recording the first baseline levels throughout the US seas. And then, from those baselines, you can then assess longer-term temporal-spatial trends and noise levels and what those patterns are. We also would like to characterize what those natural anthropogenic and biological contributions are to those different station [INAUDIBLE]. 

So in 2015, PMEL established the Ocean Noise Reference Station Network and, say, a spatially distributed network, the 12 stations throughout the US EBC. We're using the same sensor, moored at the same depths, with the same equipment to make both a fair comparison across these sites. We're targeting anthropogenic noise by staying in the lower end of the frequency ranges up to about two half kilohertz. 

And this is representing a unique NOAA-wide partnership, OAR, PMEL moving that piece, National Marine Fisheries Service. All five of the Fisheries Science Centers are involved in this work. National Ocean Service is also involved through the [INAUDIBLE] sanctuary programs. And then the National Park Service has also joined us. So just to present some of our initial results from this, I'm going to show you power spectral density plots from some of our initial data. So along the horizontal axis is the frequency range in logspace. The vertical axis is the receiving amplitude level of the recordings also in logspace. And the color represent the different station locations. 

So a couple of things to point out. Number one, the orange curve is in the Gulf of Mexico. [INAUDIBLE]. Yeah, so up here is the Gulf of Mexico, an example of anthropogenically dominated area. It's consistently and persistently higher than all the other stations. In contrast, the Arctic station, the Alaskan Arctic Station, persists below this recorded area. And the difference between these is pretty significant. It's around 30 dB, which is about 1,000 times louder than the Gulf of Mexico-- so significantly different. 

And what Carol presented earlier with retreating sea ice-- we have the baseline to make a measurement. The shipping lanes open up, and we see a lot more shipping in the Arctic, and [INAUDIBLE] we'll be able to make that quantitative measurement of what that impact is. Also here I'd like to point out the spectral peaks here that are occurring in the different stations. This is from about three months of data in the fall where you can see here we're able to identify periods where blue and fin whales are vocalizing. And that kind of gives us a seasonal habitat [INAUDIBLE] stations through time. 

So for our future research, we like to sustain the recordings of the network, be able to make those robust estimates and long-term trends both in time and space across US [INAUDIBLE]. We also want to get deeper into the site characterizations. So what are the competing roles of the anthropogenic natural and biologic sources, and also start to characterize more in-depth the habitat use by the different species in this frequency range. And hopefully, start to adopt some machine learning techniques to automate this analysis. Thanks. 

[APPLAUSE]