El Niño–Southern Oscillation (ENSO) is considered a canonical example of low-dimensional climate chaos: irregular oscillations with a 2–7 year period, asymmetry in the amplitudes of warm and cold phases, and sensitivity to random perturbations. Such systems are usually shown with graphs, but graphs require visual literacy and engage only one perceptual channel. Sonification, the translation of data into non-speech sound, leverages the auditory system's acute sensitivity to temporal structure and recurrence. However, the act of sonification itself is a nonlinear mapping between the state space of the original system and an acoustic parameter space. How well such a mapping preserves the dynamical structure of the source remains an open question. This work stands apart: nearby in the archive are studies of gravitational waves, neural network signal analysis, and phase transitions, but direct neighbors on climate sonification are almost nonexistent.
The authors took the monthly sea surface temperature anomaly index for the Niño 3.4 region from 1870 to 2024 (1860 observations). They mapped anomaly values ranging from −3 to +3 °C into MIDI notes using two Javanese pentatonic scales: pelog with intervals {0,1,3,7,8} semitones and slendro with {0,2,3,7,9}. Each note was assigned a loudness depending on both the anomaly magnitude and its rate of change. The data were then decomposed into four time scales using moving averages with windows of 3, 12, 24, and 36 months. These scales formed the basis of four compositional strategies: layered (all scales sound simultaneously), alternating (scales cycle through), melodic (12-month average leads, 24-month accompanies), and spectral (five dominant frequency peaks control timbre). The resulting audio files were segmented into frames, from which spectral centroid (brightness) and root-mean-square energy (intensity) were extracted, and trajectories were built in the two-dimensional phase space of these features.
The eight sonification variants showed systematic differences. Mean brightness ranged from 809 Hz for pelog-melodic to 1863 Hz for slendro-layered. The slendro family produced higher brightness values across all modes, consistent with its more even interval distribution. Alternating modes achieved the highest trajectory recurrence: 0.240 for pelog and 0.230 for slendro. Layered modes, in contrast, showed the lowest recurrence (0.119 and 0.123) but occupied the largest convex hull area in phase space. The correlation between brightness and energy was strongly negative for pelog (mean −0.558, down to −0.790 for alternating) and nearly zero for slendro (−0.074), with layered slendro yielding a positive correlation of +0.317. This bifurcation in coupling regimes is an emergent property of the interaction between the scale's interval structure and the mapping algorithm.
This work differs from its archive neighbors, which investigate gravitational waves, neural network filters, or phase transitions in physical systems. Here, for the first time, nonlinear dynamics tools are applied to culturally specific sonification of climate data. The choice of musical scale is shown to be not merely an aesthetic decision but a structural constraint that alters the dynamical properties of the acoustic output. Phase metrics—convex hull area, path length, exploration index, recurrence, and brightness–energy correlation—form a geometric toolkit applicable to any sonification. Distinguishing between high-recurrence, low-area trajectories (alternating modes) and low-recurrence, high-area ones (layered modes) offers a principled way to choose a compositional strategy based on the desired dynamical property of the sound.
The topic could advance toward information-theoretic measures of mapping fidelity, such as mutual information or transfer entropy between the input signal and the acoustic output. Full recurrence analysis with systematic embedding parameter selection would provide a richer characterization than simplified recurrence. Perceptual experiments with listeners, including from Indonesian communities, are needed to test communicative effectiveness. Extending to multivariate ENSO representations—sea surface temperature, atmospheric pressure, thermocline depth—could leverage the gamelan's multi-layered instrumental architecture.
The results may benefit climate communication, music informatics, and complex systems research, where rigorous methods for comparing auditory data representations are needed.
Immediate next steps include estimating mutual information between the original series and the sonification, performing full recurrence analysis with threshold optimization, and conducting perceptual tests with diverse audiences.
This work addresses the unresolved question of which dynamical properties of a complex system survive a nonlinear transformation into another modality. In climate science, quantifying information loss during time series compression and discretization remains an open problem. Sonification as a method for representing climate data lacks standardized quality metrics, and this study contributes to their development.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
In this study, a 155-year record of Pacific Ocean temperature anomalies is transformed into an eight-voice gamelan composition using two pentatonic scales (pelog and slendro) and four compositional strategies. Phase-space trajectories are constructed in the coordinates of brightness and energy, and nonlinear dynamics metrics are applied. The results show that alternating modes yield the highest recurrence of trajectories, and that the scales link brightness and energy in different ways.
Where its strength is
The strength of the work lies in combining music theory with quantitative dynamical analysis: the authors do not merely sonify data, but offer a geometric toolkit for comparing sonifications from the perspective of complex systems theory. This paves the way for an objective assessment of how well a sonification preserves the structure of the original signal.
Where it could go next
Worth trying
Why this may matter
If sonification dynamics metrics turn out to correlate with perception and understanding of data, this could provide a new tool for science communication and complex systems analysis, allowing hearing to serve as an additional channel for spotting patterns.
Nearby works in our archive
The archive is sparse around this work: the closest review is only 0.57, against a usual 0.64. It looks like few have drilled nearby.
Gamelan ensembles are not tuned to a single standard: each orchestra has its own microtonal tuning, called 'embet', making its sound unique. This study used an approximation of equal temperament, so a real Javanese gamelan would sound different.