Snowflakes shape characterization via bayesian inference: exploring the challenges

This work explores the challenges underlying the inference of characteristic parameters describing the aerodynamics of blowing snowflakes, with application to in-flight snow accretion and engine ingestion. Due to the shortage of experimental observations, the classical Bayesian problem is formulated with respect to synthetic data generated from statistics of falling snow measurements. The goal is to expose issues possibly hindering the aerodynamic shape inference process, in order to anticipate barriers and envisage solutions to apply when a comprehensive experimental data set will be available. This paper provides guidelines for implementing novel experiments, including specifications concerning the desirable accuracy and precision of measurement systems.