Impacts of eddy-mean flow decomposition on Lagrangian statistics in a barotropic double-gyre model

Synthetic particles are advected by a high-resolution flow output from a wind-driven barotropic double-gyre simulation using MITgcm, in order to investigate the impact of eddy-mean flow decomposition on Lagrangian statistics. The mean flow is usually defined as a temporal-averaged flow over the entire period of integration or observation whereas the eddy part is the remnant. Here such definition is generalized by using a spatial or temporal low-pass filter. The mean flow is defined as the low-passed part and the eddy flow is the corresponding remnant with respect to the mean part. Different eddy-mean flow decompositions yield different eddy flows. Then Lagrangian statistics derived from different eddy flows are presented and discussed. It is found that some of the statistics rely highly on the definitions of eddy-mean flow decompositions (or scale separations).