|dc.description.abstract||Waves and tides are the primary processes responsible for sediment entrainment on estuarine intertidal flats, however a detailed understanding of the temporal and spatial variability of wave processes, and the way in which these control estuarine sediment transport, is lacking. The aim of this thesis is to redress this deficiency.
Wave, current and suspended-sediment meters were deployed along a 1600-m-long cross-shore transect of the Wiroa Island intertidal flat (Manukau Harbour, New Zealand) for a period of ten weeks. Additional data were obtained by four further experiments: a pilot experiment in the shallow estuarine fringes; field experiments to test an antifouling device and to develop a way of correcting data from biofouled optical backscatter sensors; field trials of pressure and velocity sensors under estuarine and open-coast waves; and experiments to test whether optical and acoustic backscatter sensors could be used to separately measure the concentration of suspended silts and suspended sands.
Comparison of pressure and velocity data revealed an inconsistency in some wave data from the Wiroa flat and in Green and MacDonald's (2001) data from the Okura estuary: wave amplitudes estimated from pressure were significantly smaller (factor of 2-50) than corresponding amplitudes estimated from velocity data, and there was a phase lag (45-90°) between the pressure and velocity signals. The inconsistency could be interpreted as evidence of standing waves, however, data from some instruments do not support this. Tests of linear theory and sensor performance in estuarine and open-coast settings, the effect of sensor orientation, and comparisons with observed sand entrainment, suggest that the inconsistency is due to imperfect pressure-sensor frequency response.
Waves vary systematically in response to changes in fetch, which in turn are associated with the submergence and emergence of intertidal banks during the rise and fall of the tide. For a given wind, growth in fetch on a rising tide equates to growth in wave height and period. Variation in the bed-orbital speed, Uw,bed, is due to different combinations of wave height (H), period (T) and depth (h̄).When the depth is large, bed-orbital speeds are small due to poor penetration of wave motions through the water column. On the other hand, when depth is smaller, or when period is longer, penetration is more 111 effective and bed-orbital speeds are larger as a result. Consequently, in some instances, Uw,bed is largest at high tide (when fetch, H and T are large) and smaller near the start and finish of the inundation. At other times, Uw,bed is smallest around high tide (when h̄ is too large for wave orbitals to penetrate effectively to the bed). Depth-limited wave breaking was not observed, but whitecaps were common.
Sand and silt suspensions were switched on and off by wave activity (when the purewave
skin friction, θ’w , exceeded the sand threshold, θcr.sand, but not by tidal currents. Silt held in the local bed was released when sand was entrained, but there was no obvious relationship between the silt reference concentration and θ’w because a component of the silt was advected from elsewhere in the estuary and silt in the intertidal flat is supply-limited (<2% abundance). Sand reference concentration, Co, plots in three distinct clusters against θ’w which presumably correspond to rippled, transitional and flat beds. The three clusters collapsed onto a single line when flow contraction over ripple crests was accounted for, as observed by Nielsen (1986). The transitional data have not been previously identified and may represent the co-existence of rippled and flat beds. Suspended-silt concentrations were vertically homogeneous, but suspended-sand concentrations follow the simple exponential profile model (Nielsen, 1984), where the mixing length is controlled by ripple height, settling velocity and bed-orbital speed.
Silt concentrations were highest in the edge of estuarine water body known as the turbid fringe (Green et al., 1997). Enhanced settling and onshore-directed wave-orbital asymmetry are thought to aid silt deposition observed on the upper flats. Patterns in sand transport are complex because the primary drivers - tidal currents and the wave induced concentration field - are independent of each other and both vary with location across the flat and in time through the tidal cycle. This was explored by schematic modelling, which showed that combinations of different tidal regimes and fetch types produce different spatial and temporal patterns in current and concentration fields, and therefore in sediment transport. The field data showed drainage currents can be responsible for offshore transport, while asymmetry in wave-orbital motions was a means for onshore sand transport.
Spatial patterns in processes align with sedimentary and geomorphic features. Silts were found to be abundant on the channel margins, where θ’w < Ocr.sand. The intermittently inundated region near the top of the flat has the lowest slopes, and is a potential deposition zone for sands transported to the tidal edge and stranded there during the receding ebb tide. Patterns in time-integrated work due to waves, Ww, also correspond to changes in slope: slope was lowest on the upper flat where Ww was greatest, but steepened where Ww declined as a result of high tidal-translation rates across the middle flats.
Allen's ( 1971) model describes spatial trends in waves and tides across intertidal flats; wave activity at the bed increases at higher elevations while tidal currents decrease. The Wiroa data reveal a more complex situation: Uw,bed and Ww are controlled by variations in wind speed and direction, fetch length, wave height, wave period, penetration, and the tidal translation rate. Thus, while Allen established a paradigm that implies zonation in transport processes, spatial and temporal interactions between waves and tidal currents can create both more subtle and more profoundly different patterns.||