Application of stochastic data assimilation to a non-reacting shock tube configuration

An ensemble-based data assimilation method is presented for uncertainty reduction in a stochastically represented non-reacting shock tube calculation. Two modifications on the traditional ensemble Kalman filter (EnKF) are used. The first, a normal-score ensemble Kalman filter (NS-EnKF), is used to assimilate non-Gaussian distributed quantities, and the second, a feature-informed ensemble Kalman filter (FI-EnKF), enables tracking of global features. The methods are applied independently and in sequence to assimilate temporally evolving fluid states with noisy pressure observations, improving estimates of the true system state. Results are compared to the traditional EnKF and experimental shock tube data.