It has been standard practice in oil spill risk assessment modelling to source winds from the nearest meteorological station and apply these data as if they applied to the entire region to be modelled.  GEMS carried out this type of investigation into the risks associated with oil spills from the Enfield site, off Northwest Cape in November 2000.  The results of this study indicated some potential risk to the Ningaloo Marine Park during the winter months.  In this study winds from the Bureau of Meteorology station at Thevenard Island were used whereas, in a similar study for BHP at a nearby site, winds from Barrow Island were used and somewhat different results were obtained.   

Comparison of the predicted path (magenta) with the actual path (red) for the first 10 days after the drifting buoy release by Woodside Energy at the Enfield site on October 25, 2001.                            Markers are 2 days apart.

Woodside deployed a meteorological buoy in the vicinity of the Enfield site in November 2000 and asked GEMS to conduct a further oil spill risk assessment study using these data.  The results of this work exhibited significant differences to the study using Thevenard Island winds. These results raise the question as to how representative coastal wind observations are for the Enfield region and underlined the need to establish more representative wind fields for risk assessment analyses. 

To investigate these issues concerning wind forcing, and for verification of ocean current predictions with GCOM3D, Woodside released two Argos satellite tracked drifting buoys at an interval of 24 hours in October 2001.  A third buoy was released on June 28, 2002 and a fourth buoy on July 29, 2002.  

GCOM3D was run, driven by tides, winds (various sources) and geostrophic currents derived from CSIRO satellite sea surface height data and the predictions of particle drift movement from GCOM3D were then compared with the satellite tracking observations.

The results demonstrated that:

• Large errors can occur if a coastal wind site is chosen to represent offshore winds;

• Even if wind measurements from the offshore site are used it is shown that significant errors can still occur due to the lack of spatial variation in the wind fields;

• That a more realistic wind field can be obtained by using the Bureau of Meteorology forecast model winds (LAPS – Limited Area Prediction System).

• Using LAPS winds, very good agreement between the predictions of GCOM3D and the current meter measurements at the Enfield site was obtained (correlations of 0.81E/W and 0.87 N/S)

• Good agreement between the predicted and the actual tracks for three of the four drifting buoy releases.  The errors which were observed appeared to be mainly attributable to short term errors in the LAPS wind fields.

The results of the study have the following implications for oil spill risk assessment modelling (and other forms of discharge modelling): 

• The best meteorological data can be derived from the Bureau of Meteorology LAPS model with an adjustment to the wind speeds for a consistently low bias.

• Use of single station data to represent the wind fields across the model region can lead to significant errors.

• An approximation to the geostrophic currents can be obtained from satellite data.

• The tidal and wind-driven currents can be well simulated with GCOM3D.

Comparison of the predicted path (black) with the actual path (yellow) for the first 8 days after the drifting buoy  release by Woodside Energy at the Enfield site on June 28, 2002. Markers are 2 days apart