Highest Astronomical Tide for estuaries adjacent to the Great Barrier Reef

Full description

Hydrodynamic model simulations using an estuary resolving unstructured mesh to estimate the Highest Astronomical Tide (HAT) for all estuaries of the Queensland coast from Cape York to Gold Coast.
Lineage: Version 1: Highest Astronomical Tide was estimated using the Semi-implicit Cross-scale Hydroscience Integrated System Model (SCHISM) (Zhang et al. 2016). CSIRO developed an estuary resolving unstructured mesh by combining water observations from multispectral imaging satellites (https://doi.org/10.26186/146552) with a 5 m resolution composite topography-bathymetry elevation model derived from available GA ELVIS LiDAR point cloud data (https://elevation.fsdf.org.au/), DEA Intertidal (https://doi.org/10.26186/149403) and GA AusBathyTopo (GBR) 30m (https://pid.geoscience.gov.au/dataset/ga/115066). Through vectorisation of the Normalised Difference Water Index (NDWI), tidal creek allometric relationships were used to inscribe a creek depth based on a power-law function of the local creek width. This allows for tidal conveyance up into the associated floodplains. The model mesh resolution varied from 5000 m offshore down to 15 m in tidal channels and 20 m at the upper limit of the floodplain.

The model was forced with the FES2022 global tidal model (https://doi.org/10.24400/527896/A01-2024.004) on the boundary with 32 harmonic constituents. The model DEM was reduced to the Australian Vertical Working Surface quasi-geoid vertical datum (EPSG:9458) with a coastally corrected Mean Dynamic Topography (https://doi.org/10.25919/6672-jx11) applied on the model boundary to establish Mean Sea Level. The timing of Highest Astronomical Tide was estimated at 5 km intervals along the coastline using 19-year tidal predictions (2010-2029) from the FES2022 model. The unstructured mesh model domain was partitioned into subdomains with consistent tidal dynamics and simulated for each of the spring-neap cycles for which HAT occurred within the domain. The highest water level at each model point across the simulations was taken as the HAT.

The model was validated against water level observations from 30 tide gauges across Queensland including publicly available data from Marine Safety Queensland (https://www.data.qld.gov.au/dataset?q=tide+gauge) and CSIRO deployed water level radars in the Fitzroy Estuary and Cassady Creek (Hinchinbrook) - see (https://doi.org/10.25919/dxx2-s727).

After calibration of the mesh resolution and bottom friction the 2D barotropic model reproduced observed mean sea level (R^2=0.75, RMSE=0.08 m), tidal variance (R^2=0.99, RMSE=0.06 m) and Highest Astronomical Tide (HAT) (R^2=0.95, RMSE=0.16 m) well and demonstrated expected trends in tidal amplification and attenuation across the range of estuary morphologies present across Great Barrier Reef catchments.

Sub-grid hydraulic controls and infrastructure (e.g., buildings, walls, minor channels) are not explicitly modelled. Their influence on flow paths and water levels is approximated using generalized methods based on ground elevation, which may not capture site-specific behaviour.

The data, models, and analyses in this study are for informational purposes only and should not be used for local decision-making. No warranties, expressed or implied, are provided, including those of merchantability or fitness for a particular purpose.