Water Cycle Music

Since the creation of the world's first-known line chart in 1786, scientific data has been overwhelmingly represented and disseminated to the public in visual formats. Tables, charts, and graphs fill up scientific papers and PowerPoint presentations across the world, and more often then not, these representations offer little in the way of artistic expression but rather present scientific data and findings in a rather bland manner. The arts and humanities, which have been closely interwoven with the sciences since the times of the ancient Greeks, have been almost completely ignored in modern scientific communications. While science can give us a powerful understanding of how the world works, the arts and humanities are also needed to more meaningfully convey and communicate those findings to the general public.
Communicating scientific facts and data through music is one exciting possibility for the merging of the arts and sciences. To do that, though, scientific data have to be not only represented in a manner that is true to the observations that researchers have collected, but also easily interpreted by a listener and heard by a listener as a pleasing and interesting musical piece. Combining these three objectives was my challenge when, in June 2016, I arrived at the Hubbard Brook Experimental Forest in Woodstock, New Hampshire, as part of a Research Experience for Undergraduates program that was investigating ways to represent data as sound.
Coming from my third year at Oberlin College and Conservatory, where I had been studying horn performance, music composition, and mathematics, I was joining a team of scientists and musicians at Hubbard Brook that had already begun fusing science and art through Hubbard Brook's Waterviz project. Hubbard Brook scientists had been working with visual artist Xavier Cortada and musician Marty Quinn to create both realtime aural and visually artistic representations of water cycle data as well as aural representations of pre-existing water cycle data going back to 2010.
I was tasked with working with Hubbard Brook's water cycle data set from 2015. This set had 8670 data points taken at hourly intervals for ten different variables, and my charge was to make my own unique, aural representation of these data. I worked with the computer program Max/MSP, a visual programming language that allowed me to scale and store the values for each variable. The scaled values of each variable then could be assigned to an aspect of a sound or an instrument, for example pitch or volume.
To represent streamflow, I used the notation program Finale to write and record a tranquil clarinet melody. The streamflow rate controlled the clarinet's volume. If the stream flowed fast enough, cymbals would also start playing, hitting more frequently with higher streamflow.
I wrote and recorded a more spritely countermelody for the flute whose volume is controlled by solar radiation.
Solar Radiation
I used a vibraphone to represent rain and a celesta to represent snow, with heavier precipitation triggering more and higher pitches being played by those instruments.
For wind speed, I created a droning sine wave that sounded louder and played at higher octaves with higher wind speed. The panning of this drone was controlled by wind direction.
I also chose a sine wave to represent temperature. It oscillated up and down with temperature changes.
Air Temperature
I used a rolling timpani, or kettledrum, to represent air pressure, with its pitch going up and down along with the pressure.
A pan flute chord represented soil water storage levels, adding higher pitches to the chord as the soils became more saturated.
Soil Water Storage
A piano played a chromatic scale that went up and down with changes in relative humidity.
Relative Humidity
A synthesizer represented evapotranspiration, the water vapor evaporated into the air from the soil and transpired into the air by trees. This synthesizer sounds like something breathing, mimicking the trees and soils "respiring" water vapor into the atmosphere.

The end result was a fourteen-and-a-half minute piece that plays the entire year's data, with a new hourly data point heard once every 100 milliseconds. I synced the recordings up to a moving graph of the data, so that an audience can experience the data both in this new method of aural data representation and a more traditional graphical representation.

Full year

And finally, I made recordings of the following hydrological events:

Heavy Rainstorm

High Streamflow

Low Streamflow

I hope that my aural representation of data and others like it can help increase the number of people in the general public who are accessing and interacting with scientific data. The current prevailing methods of data representation (charts and graphs) make it very difficult for visually-impaired people in particular to consume scientific data. The ability to represent data both visually and aurally also provides a greater variety of ways for sighted people to interact with data. I anticipate that sonic representations presented in tandem with graphs will act as a bridge for the scientific community to this newer method of data representation.
Hopefully, more data sets will be presented in the future in more aesthetically pleasing ways, as opposed to seemingly cumbersome lists of numbers or unoriginal graphs. This will encourage members of the non-scientific community to spend more time interacting with the data, which can lead them to develop their own conclusions about it. The collaboration of artists and scientists in communicating scientific findings can allow for the exciting prospect of a world where the humanities and sciences are closely intertwined and working together for everyone's benefit.