Bed stresses in the bottom boundary layer of the York River estuary, Va., w
ere estimated from 3D near-bottom velocities measured by Acoustic Doppler V
elocimeters (ADVs) and also by a profiling array of electromagnetic current
meters. By assuming the measurements were made in a constant stress layer,
four methods of stress estimation were evaluated using ADVs: (1) direct co
variance (COV) measurement; (2) turbulent kinetic energy; (3) inertial diss
ipation utilizing the Kolmogorov spectrum; and (4) log profile. The four me
thods yielded similar estimates of frictional velocity U-* based on ADV out
put from both 14 and 44 cm above bed. All eight estimates of average U-* we
re consistent with the overall mean of 1.10 cm/s to within the 95% confiden
ce interval for individual burst estimates. The COV method worked slightly
better nearer the bed, possibly because of the sensitivity of COV to the up
per Limit of the constant stress layer. The inertial dissipation method per
formed marginally well at 14 cm above bed, likely due to sediment induced s
tratification and insufficient separation of turbulent production and dissi
pation scales. The log profile method was the most variable and appeared mo
st sensitive to stratification and to the thickness of the constant stress
layer. The turbulent kinetic energy method was the most consistent at both
heights and appears most promising for further development. Results encoura
ge future use of the ADV in estuarine environments but also favor the simul
taneous use of several methods to estimate bottom stress.