Predicted Seabed Gravel Content Grid for Darwin Harbour Region

2011-2017 survey data underpins a 10-meter resolution raster grid predicting seabed gravel content for Darwin and Bynoe Harbour. The model is based on 395 samples and 14 environmental variables, achieving a predictive accuracy improvement of 68.4% over common methods.

Use Cases

• Model spatial distribution of seabed %gravel content using the 10-meter resolution raster grid for habitat mapping.
• Analyze relationships between predicted gravel content and the 14 environmental variables derived from multibeam sonar data.
• Validate and compare gravel predictions (VEcv = 59%) against in-situ samples for marine resource management.

Strengths

• Predictive accuracy for gravel content improved by 68.4% over the commonly used IDW method.
• Model incorporates data from 395 physical samples collected between 2011 and 2017.
• Spatially continuous predictions are provided at a high 10-meter resolution.

Limitations

• Absences in predictions occur due to non-availability of associated predictive variables.
• The dataset's spatial coverage is limited to the Darwin and Bynoe Harbour regions.

Provenance

Source

Geoscience Australia, Australian Institute of Marine Science, Northern Territory Government's Department of Environment and Natural Resources.

Collection Method

Predictions derived from a compilation of multiple seabed surveys and 14 environmental variables from multibeam sonar bathymetry and backscatter.

Time Range

2011-2017

Freshness

null

Geography

Greater Darwin Harbour and Bynoe Harbour region, northern Australian marine margin.