Reynolds stress observations in continental shelf seas

The three basic observational approaches to estimating turbulence parameters in continental shelf seas are free-fall micro-structure probes from which dissipation is inferred; fast-sample (10–20 Hz) current meters in sea-bed frames measuring turbulence intensity directly; and fast-sample O(1 Hz) high-frequency Acoustic Doppler Current Profilers (ADCPs) where Reynolds stress profiles and hence turbulence production can be estimated from the variance of the along beam data. Uncertainties are associated with each approach since turbulence is a small-scale, high-frequency phenomenon and since estimates can easily be contaminated by the presence ofsurf ace waves. This paper concentrates on the latter two approaches, particularly the ADCP method, focussing on the degree of confidence that can be placed on the estimates. Results are presented from nine experiments from six sites in the North and Irish Seas and one in the Gulf of California, involving the deployment of 0.6 and 1.2MHz standard broadband ADCPs mounted in sea-bed frames. The sites ranged from very tidally energetic, shallow (20m deep) to low tidal energy, deeper (110 m). The ADCPs recorded data with a variety of sample regimes, from 2 to 0.5 Hz; bin sizes ranged from 0.25 to 1m. In two of the experiments the ADCP near-bed Reynolds stress estimates were tested against independent estimatesfrom toroidal electro-magnetic current meters measuring the three components ofcurrent (vertical and both horizontal) at 8 Hz, deployed on a nearby frame. In all cases the correlation coefficient squared between the two sets of Reynolds stress estimates was 0.7. In a further three recent deployments, an Acoustic Doppler Velocimeter (ADV) was deployed on the bottom frame with the ADV measuring volume located within the first ADCP bin and sampling at 20 or 25 Hz. The ADV measurements also show an explained variance of about 80% and a transfer function of about 1 during periods where waves were not present. One objective of these studies was to test and improve representation of dissipation processes in two- and three-dimensional numerical models, including the concept of the constant stress layer. At its very simplest, bottom stress is estimated from the depth-averaged flow via a quadratic drag law. Calculations from these measurements give values for the drag coefficient between 0.0006 and 0.0019, averaging 0.0011, smaller than the value used in most depth-averaged numerical models (0.0025). There is some evidence that the value of the drag coefficient is dependent on the tidal current speed