WAMSI PhD Top-up - 3.9.7 - Hydrodynamic processes in the Ningaloo reef system over a range of space and time scales

Brief description

As part of the WAMSI PhD Top-up scholarship - the hydrodynamic processes in the Ningaloo reef system over a range of space and time scales was studied.

A coupled wave-circulation numerical model of Ningaloo Marine Park was created. It used an extensive field data set collected from April - May 2006 in an ~5 km region around Sandy Bay, to validate its performance. The analysis of field data collected on the forereef, reef flat and in the channel revealed a strong correlation between the incident wave height and currents inside the reef lagoon and channel.

Parameters include wave parameters (wave height, period, direction) and depth average current velocity (ubar and vbar).

Lineage

Statement: An intensive 6-week field experiment was conducted during April and May 2006 when 21 moored instruments were deployed at sites spanning from the fore-reef slope to the lagoon (see WAMSI Node 3.1.6). We measured current profiles on the forereef, channel and reef flat as well as hourly directional wave spectrum. We also sampled pressure and current velocities at a fixed height typically near the middle of the water column. Tide gauges were deployed in the lagoon adjacent to the shore, and sampled water level from recorded pressure. A series of thermistor-chains were also deployed in the lagoon adjacent to the channel, to investigate vertical density stratification in the deeper regions (note that the system receives effectively no freshwater discharge from its arid coastline). Analysis of the thermistor records revealed the lagoon was well-mixed throughout the experiment (not shown), so buoyancy effects were not considered in any subsequent analysis.

The numerical modelling was based on recent nearshore developments (version 3) within the Regional Ocean Modeling System (ROMS). In particular, coupled wave-circulation modelling was conducted using the wave-current interaction routines implemented into the source code by Warner et al. 2008. The transformation of random, short-crested surface waves was simulated using the SWAN wave model (Booij et al. 1999). Wave forces, responsible for driving mean currents in ROMS, were provided by the passing of radiation stress gradients (due to wave dissipation) derived from SWAN, based on the wave-current interaction theory by Mellor (2003, 2005). Other wave-current interaction processes are also incorporated in ROMS, such as the wave-enhancement to bottom stresses and turbulent mixing. With its two-way coupling, ROMS, in turn, provides SWAN with water elevations and current fields (i.e. to account for possible current-induced wave refraction).

The hindcast simulations described above were focused on an austral autumn period of the field measurements. To investigate potential seasonal differences in the dominant circulation of Ningaloo Reef, the model was subsequently applied to additional summer and winter periods. To first understand the seasonal variability in wave and wind forcing at Ningaloo, a wave and wind climatology was initially constructed from wind observations and Indian Ocean wave model. Wave heights reach their largest values (~1.5 m) during late-winter and early-spring months (July–November), whereas a minimum (~1 m) occurs during spring (April-May). Wind speeds are greatest each year (~5.5 m s-1) during summer months (Oct-Jan) and reach a minimum during autumn (April-May).

A variety of derived hydrodynamic parameters can be defined to estimate the rate at which a coastal system, such as Ningaloo Reef, exchanges water with the ocean. We estimate the flushing time Tf of the lagoon region as (e.g., Fischer et al. 1979):

(1) where V is the volume of water in the lagoon and QL represents the flow (in m3/s) of water from the lagoon to the ocean.

The seasonal variability in the flushing time (Tf) for this section of Ningaloo Reef was thus investigated in a series of simulations. It consisted of twelve monthly simulations forced by the historical monthly-mean wave and wind conditions. To account for the mean effect of the tides (dominantly semi-diurnal), these simulations were also forced by a constant representative semi-diurnal (12 hour) tide with range 0.86 m (based on the historical mean tidal range at the site, i.e. averaging over a spring-neap cycle).

The seasonal variability in the flushing time (Tf) for this section of Ningaloo Reef was thus investigated in two sets of simulations. The first set consisted of twelve monthly simulations forced by the historical monthly-mean wave and wind conditions. To account for the mean effect of the tides (dominantly semi-diurnal), these simulations were also forced by a constant representative semi-diurnal (12 hour) tide with range 0.86 m (based on the historical mean tidal range at the site, i.e. averaging over a spring-neap cycle). The flushing time Tf was thus computed based on the velocity fields over the previous 24-hour period of the simulations (i.e. with Tf averaged over two tidal cycles).

Finally, to investigate the response of Tf to natural variability in wave and wind conditions in different seasons, two additional month-long hindcast simulations were included, focusing on a period in summer (Nov 2006) and winter (Jul 2006). These were forced by realistic time-series of wave conditions (predicted from WW3), winds (from the Milyering weather station), and tidal elevations (predicted from the OSU TPXO7.2 tide model). For the Jul 2006 period, wave conditions averaged Hs=1.5 m, Tp=15 s and the wind speeds averaged 4.1 m s-1. For the Nov 2006, wave conditions averaged Hs=1.3 m, Tp=13 s and the wind speeds averaged 5.5 m s-1.