A modified method for estimating vertical profiles of turbulent dissipation rate using density inversions in the Kuril Straits

Abstract

To address the lack of directly measured turbulence data in the Kuril Straits, an existing method was modified to indirectly estimate continuous vertical profiles of turbulent energy dissipation rate e by using density inversions. A linear relationship was confirmed between directly measured e and indirectly estimated e from the existing method, where most of the detected density inversions were discarded as noise. The existing method thus yielded large gaps in the vertical profiles, and the gaps were much greater than the observed mean autocorrelation vertical length scale of about 10 m. To reduce these gaps and produce reasonable estimates for vertical e profiles, the quality tests were carefully re-examined with directly measured e data, and one of the quality tests (the water mass test) was excluded because the test rejects real density inversions even with large e. With this modification, and interpolation and smoothing of the indirectly estimated e with the mean correlation length scale, continuous vertical e profiles were obtained. These profiles have an error factor of 3.3 corresponding to one standard deviation of the ratio between directly observed and estimated data, and 95 % of the data were within a factor of 10.5, with the overall correlation coefficient between smoothed directly measured e and estimated e equal to 0.80. This method could be useful for areas where 10−9 < e < 10−6.5 W/kg, and where vertical distances between adjacent density inversions are mostly less than the mean autocorrelation scale.