Bottom boundary-layer observations beneath NLIW of depression; April 3

Observations made in the bottom boundary-layer beneath a large amplitude NLIW of depression on April 3. The observations have been quality controlled and corrected for motion where necessary. The data have been used in a manuscript titled "Observations of Enhanced Sediment Transport by Nonlinear Internal Waves" submitted to Geophysical Research Letters.

The two ADV data sets are derived from Nortek Vector ADVs. They have been cleaned using the standard Nortek Vector data quality diagnostic metrics and despiked following the method of Goring and Nikora (2002). The data have been rotated onto a fixed geographical frame of reference. The mean component of the vertical velocity is to be treated with suspicion. Very slight errors in tilt sensors lead to contamination by the mean horizontal velocity, which is orders of magnitude larger than the mean vertical velocity. This is a known challenge of near-bed vertical velocity measurement, and these mean velocities are not used in the manuscript. The fluctuating components of vertical and horizontal velocity are more similar in magnitude, reducing the degree of contamination. Fluctuating vertical velocities have been sense checked by spectral scaling, non-dimensional stress ratios and turbulence closure estimates, and they have demonstrated to be of high quality.

The ADCP data provided include the raw (8 Hz) backscatter and mean (5-minute mean) velocities, as used in the manuscript. Prior to averaging the ADCP velocities were cleaned using standard on-board diagnostic measures. A slightly modified ‘fish-detection’ algorithm was used as spatial heterogeneity in turbid ejections in the boundary layer would be flagged as spurious by standard algorithms. The vertical velocity near-bed is to be treated as suspicious. There is thin layer of constant vertical velocity near-bed which moves in height depending on the thickness of the suspended sediment layer. The velocities above 5 m are considered reliable. No spurious near-bed velocities were used in the manuscript.

The temperature data from the T-string data are calibrated using coefficients derived by a certified external laboratory, corrected for mooring motion, interpolated onto a regular grid in the vertical, and temporally averaged to the sampling frequency of the lowest frequency instrument in the array (15 s). Otherwise these are raw.

The turbidity data are calibrated using coefficients derived by a certified external laboratory. Otherwise these are raw.