Salinity effects on pressure-based tide gauges in a macro-tidal estuary

<div>Estuarine salinity can be highly variable, resulting in significant density variability. So tide-gauges measuring pressure (either directly or as bubblers) can experience salinity-induced pressure variations of up to 1% of their range. They can therefore record depths with up to 1% variability that is not seen in other measurement systems. In most places this is a small effect compared to other sources of varibility but in macro-tidal estuary such as the Severn in the UK it amounts around 10cm difference between high and low tide. This is comparable to levels of accuracy being sought by current surge forecasts. </div><div> </div><div>This effect can be seen in neighbouring radar and pressure gauges in the Severn. It could result in spurious seasonal and spring-neap constituents. Elsewhere, similar errors could result related to unusual freshwater river flow in estuarine locations that are usually saline, or unusual saline storm surge in usually freshwater sites. It has implications for assessment of storm surge and multi-hazard forecasting, and accurate calibration of new gauge installations.  </div><div> </div>

Hydrodynamic modelling and model sensitivities to bed roughness and bathymetry offset in a micro-tidal estuary

Tidal estuaries support everyday functions for over 80% of Australia's population living within 50 km of the coastline and thus come under immense pressure of physicochemical changes. Most studies in estuarine applications have used the bed roughness as the single calibration parameter to calibrate hydrodynamic modelling, yet errors in bathymetric data can significantly impose uncertainties into the model outputs. In this study, we evaluated the sensitivity of a hydrodynamic model of a micro-tidal estuary to both the bed roughness and bathymetry offset through comparing observed and modelled water level and velocity. Treating both bathymetry offset and bed roughness as calibration parameters, three calibration scenarios were tested to examine the impact of these parameters. To validate the model, Lagrangian drifter data as a new dataset in shallow estuaries were used. The analysis shows that model outputs are more sensitive to the variation of bathymetry offset than bed roughness. Results show that calibrating the bathymetry offset alone can significantly improve model performance. Simultaneous calibration of both parameters can provide further improvement, particularly for capturing the water level. Drifter and modelled velocities are highly correlated during flood tides, whereas the correlation is low for slack water because of wind-induced current on drifters.