A Modal Description of Dynamic Wake Meandering

Lidar scans from a nacelle-mounted measurement system provide time series of wake measurements during varied atmospheric inflow conditions, from which we describe the coherent turbulent structures that contribute to wake meandering through proper orthogonal decomposition. Subsets of modes are used to make low-order reconstructions in a combinatorial sense, yielding more than 30,000 estimates of meandering for each inflow case. A regression test using the range of reconstructed flow statistics identifies the modes that contribute most to wake meandering. Mode coefficient spectra highlight the dominant Strouhal number associated with each turbulent structure, suggesting that the lowest ranking modes do not necessarily contribute most to the accuracy of the reconstruction. Instead, some modes appear to have no influence on meandering dynamics, and still others consistently detract from wake meandering represented in low-dimensional flow reconstructions. No consistent relationship is revealed between characteristic frequencies for each mode and for either the inflow or the lidar measurements, suggesting that a more complex relationship between wake and inflow turbulence may be needed to accurately describe wake meandering.