Scientists are recording the sound of the whole planet

By Josh Dzieza
 

http://www.ascensionearth2012.org/2014/08/scientists-are-recording-sound-of-whole.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+AscensionEarth2012+%28Ascension+Earth+2012%29

Researchers are listening to everything from airplanes to bat calls in order to learn more about the state of the environment

In a few weeks, sensors in Indiana will go online that will record, in the words of Bryan Pijanowski, every sound the Earth makes. The array of microphones, geophones, and barometric gauges will run for a year, taping everything from the songs of birds arriving in the spring to the vibrations of the continent as ocean waves pound the Atlantic and Pacific coasts. They will measure earthquakes on the other side of the world and the stomping of cattle nearby, the ultrasonic whistles of bats and the barometric drop of cold fronts. “I joke to my physicist friends that if I had a microphone small enough, I could record the Higgs boson,” Pijanowski says.

A color-coded spectrogram of 24 hours of noise in the Australian bush, by Michael Towsey of the Queensland University of Technology. The morning chorus starts at 4:30AM and the evening cicada chorus around 6:00PM.

Pijanowski is a soundscape ecologist, a term he coined three years ago to describe a new approach to studying sound. Rather than look at how, for example, a single species of frog calls for a mate, soundscape ecologists study how all the sounds in a space interact, from frog calls to car traffic to thunder. "There are what I call rhythms of nature, there are periodicities like the dawn chorus and certain crescendos during the seasons," Pijanowski says, referring to the way birds burst into song at sunrise. He believes listening to these patterns can tell us important things about the state of the natural world.

Though still small, the field is growing, thanks in no small measure to Pijanowski’s tireless efforts (and those of his grad students). For the last several years he’s been circling the globe, depositing microphones in Costa Rica, Borneo, Tippecanoe, the Sonoran desert, Alabama, the wildfire-ravaged Chiricahuas, and urban parks in Chicago, often giving talks along the way. You get the sense he’s slightly reserved except when talking about sound, at which point he gestures expansively and uses words like "marvelous," "magnificent," and "glorious."

"There are what I call rhythms of nature."

Earlier this year, Pijanowski launched the Global Soundscape Project, which is building a map of the world's sounds using an app that turns phones into recorders. Occasionally he has an IMAX crew in tow, part of a soundscape education program he’s filming. And every few months he convenes soundscape researchers for workshops, part of a grant from the National Science Foundation. He invites people from outside the sciences to participate. "When you look at arrangements of sound, working with musicians helps you to think about the orchestration of an ecosystem," he explains.

The idea that animal sounds follow a complex order goes back to Bernie Krause, a musician who in the 1960s and ’70s made a living doing sound work for the film industry, frequently taping things like jungle noises and whale songs. He became enamored of nature sounds and started accompanying researchers into the field to make recordings, eventually becoming the preeminent wildlife acoustician. In 1985, he was called on to lure a confused humpback whale, Humphrey, out of the Sacramento river using a feeding song.

Pijanowski's recording of cicadas in Borneo

Pijanowski's recording of an hour in the La Selva rainforest in Costa Rica. The top spectrogram is the full hour and the three bottom ones are zoomed-in images of the 26 seconds following each of the three red markers. The audio corresponds to the zoomed in images.

As he sat in jungles and deserts around the world, Krause noticed that the sounds he heard could be surprisingly orderly. Different species seemed to occupy their own place in the sonic spectrum. Insects in Borneo might stridulate loudly at a middle frequency, alternating so as not to drown each other out. Birds rise above it by calling at a higher pitch, and birds with shorter calls fit in-between the calls of birds with longer ones. Frogs puncture the droning insect noise with short, loud bursts, and mammals take the bottom frequencies. Organisms, Krause hypothesized, evolved to partition acoustic bandwidth, calling out at different frequencies and at different intervals to be heard over one another. Animals would also have to evolve to be heard over sounds like thunder, wind, and rushing rivers — sounds that Krause, working with ecologist Stuart Gage, called geophony. And in more recent history, animals must also adjust to anthrophony: the sounds of human civilization.

Krause called his idea the acoustic niche hypothesis, and it had a corollary. If organisms evolved to share the acoustic spectrum, maybe disruptions from pollution, development, invasive species, and other threats would result in gaps in the arrangement of sounds. In 1989, Krause found what he believes is evidence of such audible damage. The year before, he had taken a recording of a forest in the Sierra Nevadas. He returned after it had been selectively logged and found the soundscape almost silent.

The idea that you can hear environmental damage is evocative — Rachel Carson knew that when she chose the title Silent Spring — but as powerful as Krause’s Sierra Nevada recording is, there are other potential explanations. It could have been a La Nina year, Pijanowski says, causing the birds showed up later. There could have been landscape changes elsewhere on their migratory route. There was no control group, no uncut forest in the same area to measure against.

In the last several years, researchers armed with microphones and data-sifting algorithms have been trying to explore and build on Krause’s ideas. They’re using microphones to monitor biodiversity in Costa Rica and Australia, and a similar network is being established in Germany. Other researchers have diagnosed dying coral reefs by the sound, as various fish and crustaceans go silent.

Pijanowski believes he’ll be able to hear shifts in the soundscape as the climate changes. Insects, whose life cycles are driven by temperature changes, will emerge earlier, while birds and mammals, whose behavior is driven primarily by the length of the day, will remain the same. Amphibians are driven by both factors, so it’s unclear how they’ll respond. New species will invade warming regions, potentially adding their own calls or silencing those of native animals. "We will start to hear a reassembly of the soundscape as summer comes earlier," Pijanowski says.

Amandine Gasc and Matt Harris retrieve a Songmeter recorder and two cameras

Pijanowski is responsible for much of the field’s recent growth, both by giving a name to what disparate researchers were doing, and by convening many of those researchers in workshops. The last one was held in Maine — fittingly, near the Rachel Carson Wildlife Preserve — and drew a group of ecologists, biologists, musicians, engineers, artists, and philosophers.

"It’s still in its renaissance period," explained Tom Seager, an ecologist from Arizona attending the workshop. "Where both technologists and artists can contribute."

As a fledgling field, there was a lot of debate over terms and concepts, discussions that frequently ended up in philosophical territory. One such debate was over what to call noises that humans make.

"I no longer like the term anthrophony," Stuart Gage said as he walked through the forest listening to birds. A gray-bearded, soft-spoken former entomologist-turned-soundscape ecologist, Gage has a measured way of speaking and a saintly determination to neither use insect repellent nor to swat the swarms of mosquitoes battening onto him. If Krause is the godfather of soundscape ecology and Pijanowski its current evangelist, Gage is the bridge. He helped Krause come up with the taxonomy of sound in the early 2000s and advised Pijanowski on his thesis. "I’ve argued with Bernie a number of times that we ought to use the term technophony to distinguish sounds humans make from technological sounds — because humans are critters too, we communicate in the same way, with our voices. But we also make things."

Jeff Migliozzi, a teacher at the Perkins School for the Blind, agreed. "You’re essentially redefining man, saying instead of being a biological creature, we’re creators of technology, and the rattle and the hum."

"Maybe that’s true," said Gage.

A model of different forms of sound and silence in the Kenai Wildlife Refuge by Tim Mullett, University of Alaska, Fairbanks.

Out on the estuary, Pijanowski was checking a recorder he’d set up in May. It had captured the pounding surf, shrieking gulls, sparrows, crickets, and hawks. The tides set the rhythm: high tide was silent and low tide cacophonous, as the birds swooped down to devour animals trapped in the tide pools. There was a rhythm to the technophony too, a dawn chorus of diesel engines as fishermen moved up and down the coast, weekly influxes of jets and speedboats, heavier on the weekends and increasing into summer.

The issue of mechanical noise was a major theme of the workshop, and of soundscape ecology in general. Falk Huettmann from the University of Alaska Fairbanks projected a noise map of the Kenai Wildlife Refuge made by his graduate student Tim Mullett. Mullett had traveled deep into the glacial refuge to set up microphones, going high into the mountains dozens of miles from the nearest road. He still found mechanical noise everywhere, mostly from airplanes and snowmobiles.

Speaking grimly in a German accent, Huettmann declared, "We need to abandon the idea of wilderness. It doesn’t exist."

Mechanical noise impacts different animals in different ways. In some cases — sonar and marine mammals, to name one — it’s disorienting and damaging. In others, animals adapt in ways we’re just beginning to understand. Grasshoppers that live near roads evolve to call at a higher pitch, to be heard over traffic noise. Even when taken to a silent room in a lab, they stridulate at a higher frequency than more rural grasshoppers, which makes sense — only grasshoppers that can be heard above the cars would find a mate.

"We need to abandon the idea of wilderness. It doesn’t exist."

There seems to be variation in how birds respond to noise. In 2003, researchers found that great tits (a bird) living in loud parts of the Dutch city of Leiden called at a higher pitch than tits in quiet parts. Urban robbins, meanwhile, appear to call at night not because they’re confused by city lights, as previously thought, but because they want to avoid the noisy day. Another researcher found that responses to noise varied depending on the species: one bird, the gray flycatcher, fled areas where gas drillers were using noisy air compressors, while ash-throated flycatchers simply called at a higher frequency. Sharon Gill, a biologist at Western Michigan University who was at the workshop, is studying how individual chipping sparrows respond to noise. Her initial findings indicate that there’s great variation in how individuals react, with sparrows with deeper calls raising their pitch more drastically to be heard over the sound of traffic.

"I’m really interested in the persistence of species in a changing world," Gill says. "This isn’t the environment these animals evolved in, with these high levels of noise."

The sound of nine months in the Australian bush, modeled by Michael Towsey of the Queensland University of Technology. The white lines represent dawn and dusk.

Soundscape ecology’s current challenge is finding a way to sort through the vast amounts of data being collected. In just a few years, Pijanowski’s lab has accrued tens of thousands of hours of audio.

There’s no way anyone could listen to all this audio, so algorithms need to sort through it. Current methods are somewhat crude. Gage’s index takes everything at a frequency below 2 kilohertz and labels it human, then quantifies the acoustic energy in the ranges above that. It’s roughly accurate, but certain animals, like the loon, call at a low frequency. Another method sorts sound by its shape on a spectrogram. Animal sounds are generally short and sharply peaking, whereas machines tend to drone at a constant level. Again, though, it’s not always true — think of crickets in the summer, or the aquatic bang of the air guns used in undersea oil and gas exploration. Machine learning could help correct these issues, though that research is just beginning.

Computer scientist Michael Towsey in Brisbane, Australia, is taking a visual approach, using multiple indices to create color-coded images of soundscapes. The idea is that a trained ecologist could look at the charts of sound over the course of a day, month, or year and identify changes, like an acoustic weather map.

The hope is that an algorithm with the right index will parse the audio ecologists are collecting, turning low-cost microphones into a powerful network of sensors. With a way to quickly digest audio, researchers would have a vast array of data on what species are where and when, data that over time could provide a valuable glimpse into the way the environment is changing.

"I looked for most of my scientific career for an instrument that would measure the environment," Gage says. "I use the analogy of a stethoscope. A doctor can use a stethoscope to tell 10 different things about your heart. We’re holding a stethoscope up to nature. We’re listening to the heartbeat of the environment, whether it’s the heartbeat of a city or the heartbeat of a forest, it’s the heartbeat of the biosphere."