Analysis of Flow Patterns in NSTX Edge Turbulence

The dynamics of turbulent fluctuations near the edge of plasma confinement devices are thought to have a significant impact on plasma performance. The Gas Puff Imaging (GPI) diagnostic provides movies of density and temperature fluctuations in the tokamak edge region with high temporal ($\sim$4 microsecond) and spatial ($\sim$1-2 cm) resolution. The Principal Component Analysis (PCA) method provides an optimized representation of each movie in terms of empirically derived spatial `eigenfunctions' multiplied by time-dependent amplitudes. In this poster, new methods for the inference of a velocity field from the PCA representation of turbulent fluctuations, as well as implications related to the theory of edge turbulence, are discussed.