AbstractWe use profiles from a Lagrangian float in the north Indian Ocean to explore the usefulness of Thorpe analysis methods to measure vertical scales and dissipation rates in the ocean surface boundary layer. An rms Thorpe length scale LT and an energy dissipation rate εT were computed by resorting the measured density profiles. These are compared to the mixed layer depth (MLD) computed with different density thresholds, the Monin–Obukhov (MO) length LMO computed from the ERA5 reanalysis values of wind stress, and buoyancy flux B0 and dissipation rates ε from historical microstructure data. The Thorpe length scale LT is found to accurately match MLD for small (<0.005 kg m−3) density thresholds, but not for larger thresholds, because these do not detect the warm diurnal layers. We use ξ = LT/|LMO| to classify the boundary layer turbulence during nighttime convection. In our data, 90% of points from the Bay of Bengal (Arabian Sea) satisfy ξ < 1 (1 < ξ <10), indicating that wind forcing is (both wind forcing and convection are) driving the turbulence. Over the measured range of ξ, εT decreases with decreasing ξ, i.e., more wind forcing, while ε increases, clearly showing that ε/εT decreases with increasing ξ. This is explained by a new scaling for ξ ≪ 1, εT = 1.15B0ξ0.5 compared to the historical scaling ε = 0.64B0 + 1.76ξ−1. For ξ ≪ 1 we expect ε = εT. Similar calculations may be possible using routine Argo float and ship data, allowing more detailed global measurements of εT, thereby providing large-scale tests of turbulence scaling in boundary layers.