Predicting neutrino flavor evolution in supernova astrophysics, and acoustic structure in birdsong: two unconventional applications of data assimilation
The evolution of neutrino “flavor” following a core-collapse supernova (SN) event is an important question for cosmology, because that flavor evolution in-part sets the abundances of the heavy elements in the Universe. We sought to ascertain what information an earth-based detector must receive in order to infer the flavor evolution history of neutrinos that emanated from such an event. Using a simplified model of neutrino interactions, a method of statistical data assimilation (DA) can infer the complete flavor evolution history, given measurements only at the location of the detector.
A similar DA protocol can be employed to synthesize the acoustic time series pressure wave generated by a songbird. This discovery we made with the aim of identifying a predictive metric for observed song preferences in female birds. To date, most acoustic analysis tools have failed to identify such a metric. This is not surprising, given that those tools are based on linear spectral analysis, while vocalization is a complex nonlinear phenomenon. We show how DA can reliably reconstruct the acoustic structure of song, by assuming that it was generated by a nonlinear dynamical model.
Dr. Eve Armstrong University of Pennsylvania