Shearing stresses and turbulence in a tidal current.
Simultaneous measurements of the velocity profile and turbulence were
made at 4 heights within 2 m of the sea bed in the Eastern Irish Sea. A
photographic unit was also deployed to observe sediment motion. The design,
construction and performance of the instruments are described and the
results from 36 trials, for a variety of bedforms and water depths of 8 to
50 m, are discussed.
The velocity profile was recorded by means of Aanderaa rotors and the
velocity fluctuations with e.m. heads. Their measurements, according to
flume calibrations, were accurate to ±l%. Angular corrections were
necessary to compensate for the loss in response when the turbulence rig was
poorly orientated with respect to the mean flow direction. The minimum in
situ velocity was 20 cm s-l, at which speed inertial 'pumping' of the rotors
by turbulent fluctuations was estimated to produce overreading by - 1 cm
s-l. At speeds above 30 cm s-l this effect was negligible. The resolution of
the e.m. heads was at least 5.0 mm s-l, as determined by the noise level,
and d.c. drift less than 1.0 mm s-l over periods of up to 17 hours. Data
were recorded on 9 track tape aboard ship for later analysis onshore.
The measurements were made in a bottom boundary layer which could be
decribed as accelerating, non-rotational, hydrodynamically rough, neutrally
stratified and comprised a layer of constant Reynolds stress.
The stress, as determined from the log-profiles, either uncorrected or
corrected for acceleration, was observed to be significantly greater (-
26%) than that determined by eddy correlation techniques. This could not
be attributed to uncertainties introduced by misalignment of the e .m.
heads, or inadequate correction for cospectral losses. Doubt was cast on
the validity of von Karmann's constant (KO) = 0.4, with a more appropriate
value apparently being closer to 0.35. KO and u*2/-U7""WTexhibited no
dependence on bedform or sediment type, except in one case, where high
ratios corresponded to high zo's.
Evidence of a tidal hysteresis of stress was observed at one station
only. Apparent Zo minima at peak tidal velocities were, for the most part,
attributed to the non-removal of accelerating effects when applying the
log-law. Initially high values of zo' when present, were supposed to be
due to streamlining of bedforms with ;ncrnasing velocity. Zo and C100
varied from 0.02 - 0.25 and (2 - 3) x 10-3 respectively for mud, unrippled
sand and various sand combinations. For gravel and rippled sand values
they were 0.10 - 0.20 and (3 - 8) x 10-3.
A comparison of events comprising 90% of the stress with movement of bed
material, observed by the photographic unit, failed to reveal a
Correlation. During periods of sediment motion events in which u' > 0 were
dominant. For events comprising 90% of the stress, those in which u' < 0 and
u' > 0 occurred in groups of 5 - 20. In addition, a number of coherent
events, occuring between 5 - 12 times per minute independently of velocity,
were observed between 100.0 and 172.5 cm above the bed.