2026/05/05 by Yuchen Ma, Raffaele Ferrari, Kurt Polzin +2
Earth and Planetary Sciences · Engineering · #Oceanographic and Atmospheric Processes #Fluid Dynamics and Turbulent Flows #Geological formations and processes
paper · doi:10.1175/jpo-d-25-0286.1
Abstract Ocean turbulence is highly intermittent, making statistical models crucial for interpreting sparse microstructure data and quantifying mixing rates. Classical theory predicts a lognormal distribution of oceanic turbulent dissipation, and recent work suggested that a log-skew-normal distribution better describes microstructure data. In this work, it is shown—using turbulent dissipation data from the Brazil Basin and North Atlantic tracer release experiments—that the log-skew-normal fit systematically overestimates the largest mixing events near the ocean bottom. Instead, the data are well described by a mixture of two lognormal distributions: one with a small mode representing a population of relatively weak mixing events and another with a larger mode representing a distinct population of more vigorous mixing events. The mixture model is consistent with a partitioned two-regime view of oceanic dissipation related to two dynamically different stratified turbulent processes. Several plausible (and largely consistent) hypotheses regarding the physical origins of two different pathways of mixing are discussed. Significance Statement Our study revisits the statistics of oceanic measurements of a physical quantity that characterizes the strength of ocean turbulence. We find that these statistics are bimodal, with two distinct peaks: one associated with strong turbulence events like “bursts” and the other with relatively weak background turbulence. This suggests two different pathways for energy transfer to small scales, which may substantially change how we interpret ocean turbulence.