Uncertainty Estimation for 2D PIV: An In-Depth~Comparative Analysis~

Uncertainty~quantification methods have recently~made great strides in accurately predicting uncertainties for planar PIV, and several different approaches are now documented.~ In the present study, we provide an~analysis of these methods across~different~experiments and different PIV processing codes.~ To assess the performance of said methods, we follow the approach of Sciacchitano et al. (2015) and~utilize two PIV measurement systems with overlapping fields of view---one acting as a reference~(which is validated using simultaneous LDV measurements)~and the other as a measurement system,~paying close attention to the effects of interrogation window overlap~and bias errors~on the analysis.~~A total of three experiments were performed: a jet flow and a cylinder in cross flow at two Reynolds numbers. In brief, the standard coverages (68{\%} confidence interval) ranged from approximately 65{\%}-77{\%} for PPR and MI methods, 40{\%}-50{\%} for image matching methods. We present an in-depth~survey~of both~global (e.g., coverage and error histograms) and local (e.g., spatially varying statistics) parameters to examine the strengths and~weaknesses~of each method~in monitor their responses to different regions of the experimental flows.~~ ~