The bronze plates of a gamelan do not sound according to a single tuning standard: each orchestra on Java keeps its own tuning, and two neighboring ensembles may play the same melody in slightly different intervals. This fact became the starting point for an unusual experiment: scientists translated 155 years of Pacific temperature anomalies into notes of two Javanese scales and examined how sound preserves the dynamics of climate chaos.

The climate phenomenon El Niño–Southern Oscillation (ENSO) resembles a pendulum that swings irregularly: warm phases alternate with cold ones over periods of two to seven years, with warm deviations being sharper than cold ones. Usually such oscillations are shown with graphs, but graphs engage only vision. Hearing, however, is keen at detecting recurrence and temporal structure, so the data were converted into sound, a technique called sonification. The question was: how accurately does the acoustic snapshot convey the chaotic nature of the original system, and how does the choice of musical scale affect that?

To separate fast fluctuations from slow ones, the temperature anomaly series was decomposed into four time scales using moving averages. From these scales, four compositional strategies were assembled: layered, where all scales sound simultaneously; alternating, where they succeed one another; melodic, where one scale leads and another accompanies; and spectral, where five dominant frequencies control the timbre. The resulting audio files were split into frames, two sound features were extracted, brightness (spectral centroid) and energy, and trajectories were plotted in the two-dimensional space of these features.

It turned out that the choice of scale and strategy dramatically changes the behavior of the trajectories. Alternating regimes yielded the highest recurrence: the trajectory often returned to previously visited regions, like a person walking familiar paths. Layered regimes, in contrast, scattered trajectories over a larger area but returned less often. The coupling between brightness and energy also diverged: for pelog it was strongly negative, meaning higher notes were quieter, while for slendro it was nearly zero, and layered slendro even gave a positive correlation. This split was not preordained; it emerged from the interaction between the scale's interval structure and the mapping algorithm.

The work provides a tool for comparing any sonifications of complex signals. Previously, it was unclear whether a sound representation preserves the dynamical structure of its source; now one can measure this via the area of the convex hull, path length, exploration index, and recurrence. The methodology will be useful in climate communication, where sound can convey alarming trends to those unaccustomed to reading graphs, and in music informatics, where understanding how mapping algorithms affect perception is crucial. The bronze plates of the gamelan, tuned differently in each village, turn out to be not just an exotic curiosity but a sensitive instrument for studying chaos.