The residents of the town of Nenana in Alaska have an unusual annual tradition. Every March, as the days begin to grow longer and the sun rises higher in the sky, they assemble a tripod out of local timber and carry it to the banks of the Tanana river that flows past the northern edge of the town.

In the winter, the river does not flow. With daily temperatures that fall far below zero (in both Fahrenheit and Celsius), its surface is frozen for about six months of the year. The town’s residents carry the tripod out onto the ice, carve a trough about a metre wide with a chainsaw, place the legs of the tripod into the hole, and freeze it in place. A rope is connected to a clock, inside a tower, on the shore of the river, and the “Nenana Ice Classic” begins.


For the last 18 months,  I’ve been learning how to turn data into music and sound. Alongside my collaborator, Miriam Quick, we’ve been learning the dark arts of “data sonification”–where numbers in a dataset are mapped not onto colour, length or angle, but volume, pitch, and tone.

Sonification has been around for more than a century, at least since physicist Hans Geiger invented the Geiger counter in 1907, for measuring ionizing radiation. The device emits a distinctive set of clicks that increase in number as it’s brought closer to a source of radioactivity. 

But despite its long history, despite it being around us every day, and despite it being a fantastic accessibility tool, sonification doesn’t get anywhere near as much attention as visualization in the data community. A search for “sonification” in the Data Visualization Society’s Slack instance, at the time of writing, yields just two results – one of which was posted by me. A growing sonification community exists, but it’s quite separate from the visualization community. 

Miriam and I want to change that. We believe that sonification is mature enough and powerful enough to be included in visualization experts’ day-to-day toolbox, rather than seeing it as a novelty to be explored when a client wants something a bit unusual. With mature libraries like Tone.js (for Javascript), Pya (for Python), and Sonify (for R), as well as dedicated tools like Sonic Pi, CSV to MIDI, and TwoTone.io, it’s time to usher sonification into the mainstream.


By Frank K. from Anchorage, Alaska, USA – Tripod on the Tanana River in Nenana – part of the Ice Classic to determine break-up (IMG_4558a), CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=3741189

The Nenana Ice Classic began in 1906 when Nenana locals, many of whom had arrived after gold was discovered in Fairbanks, began a betting pool on when the ice would break up in the river each year. 

The list of entries in the first competition is a who’s who of the frontier town at the time: Adolph Nelson, Jim Duke (who owned the local trading post), Louis and Joe Johnson, a guy called “Jonesy,” another guy called “Gunnysack Jack,” and the winner that year – Oliver Lee, who won enough to pay for a couple of rounds at the trading post bar.

After that, the idea was forgotten again for a few years, but in 1916, workers who were in town building a railroad from Anchorage revived the competition, selling tickets through Jim Duke’s Roadhouse. Word spread along the rail line, and in 1917 the prize pool had swelled to an impressive 801 dollars. That year the clock stopped, indicating that the ice had officially broken up, at exactly eleven thirty am on the thirtieth of April.

More than a century later, somehow, it’s still going. The contest has been held every year since, with the prize pool swelling to hundreds of thousands of dollars. People who want to enter buy tickets, which they deposit in one of the famous “red cans” found at more than 200 places all over Alaska. Some of the money goes towards funding scholarships, kids’ sports teams, and local charities.

But in recent years, people hoping to grab a chunk of the jackpot have a new factor to contend with–climate change. In 2019, the ice broke up on the fourteenth of April – the earliest date on record, and of the ten earliest breakup dates, five have occurred in the last ten years. It turns out that the Nenana Ice Classic’s “Book of Guesses” doubles as a record of how humans are changing the composition of our atmosphere. Of how our planet is warming. Of how the frozen landscapes of the polar regions are melting beyond recognition.


On 5 June 2021, we released the first episode of the first season of Loud Numbers – a data sonification podcast. The format is the same in every episode: we introduce a data story, explain how we created a sonification from it, and then play back that sonification. If you like podcasts like Song Exploder, 99% Invisible, and 20,000 Hz, you’ll like this, too.

In the first episode, we’ve turned Nenana’s “Book of Guesses” into an ice-cold techno track, combining multiple datasets into a single piece of music. It begins in 1901, because our data begins in 1917, every two bars are a year, and dance music is often grouped into multiples of four bars.

The first data layer is the solar cycle. During the winter, the aurora borealis swirls through the skies of Alaska and its strength rises and falls in eleven-year sunspot cycles. Humans have been counting sunspots for hundreds of years, tracking these cycles, so we brought them in as an ethereal shimmer in the background. These shimmers are based on real data from the Royal Observatory of Belgium – the louder the sound, the more sunspots heard in a given month.

After the introduction, when the timeline reaches 1917, we start to sonify the Nenana ice breakup dates. We used the date the tripod fell each year, took a ten-year average to smooth out the noise in the dataset, and mapped the result to the pitch of a chord. The higher the chord, the earlier the ice broke up on the Tanana river. From this point on, there’s a distinctive click sound each time a new year begins.

The tripod chords go up and down in pitch, but on the whole they get higher as the track progresses, showing the ice melting earlier and earlier as climate change in Nenana takes hold. We had some fun with the filters on these chords, but they don’t encode data – they’re just there to create hands-in-the-air moments. We’ve found it’s important to include sonic elements that aren’t tied to data alongside those that are. It makes the result sound more musical – which deepens the emotional connection with the listener.

We also wanted to sonify the reason for all that warming ice. And it’s not the sunspots. In the background of the track – too faint to hear at first, but louder and louder over time – you’ll hear a siren. The pitch of the siren represents the amount of carbon dioxide in the atmosphere, measured by the observatory at Mauna Loa in Hawaii. The CO2 level rises and falls each year as forests grow and die back in the Northern Hemisphere, which has more land. It’s like the Earth breathing. That’s why the pitch of the siren wobbles a little. But it also increases over time and rises to a terrifying climax near the end of the track.

After all that description of the sound, you probably want to hear it, right? Well you can! The podcast edition includes a description of the sonification system, complete with audio examples:

Or you can just listen to the track on its own if you prefer:

This is just the first installment of the first season of Loud Numbers – there are four more that’ll be released over the course of June and July 2021. Other episodes include music made from data on the US economy, beer tasting, EU bureaucracy, and the disappearance of insects, set to musical styles that range from UK jungle to Baroque counterpoint.

We hope you’ll join us for the journey. You can sign up for our newsletter at loudnumbers.net, and search for and subscribe to the podcast in all good streaming services.

Author profile

Duncan Geere is an information designer based in Gothenburg, Sweden, interested in climate and the environment. He works to communicate complex, nuanced information to a wider audience for clients like Information is Beautiful, the Gates Foundation, and Project Drawdown. He currently works part-time for the climate charity Possible, and he’s also a generative artist and musician.

Author profile

Miriam Quick is a data journalist, researcher and author who explores novel ways of communicating data. She has written data stories for the BBC, worked as a researcher for Information is Beautiful and the New York Times and co-created artworks that represent data through images, sculpture and sound. Her first book, I am a book. I am a portal to the universe., co-authored with Stefanie Posavec, was published in September 2020.