Bass Strait Sediment Grain Size Prediction Grids

Updated: 2026-08-10

<p>These datasets describe sediment grain size properties at the Bass Strait region, Australia. They include six variables:&nbsp;</p><ul><li>mud content (mud_BS.tif),&nbsp;</li><li>sand content (sand_BS.tif),&nbsp;</li><li>gravel content (gravel_BS.tif),&nbsp;</li><li>sediment classes (folkClass_BS.tif),&nbsp;</li><li>mean grain size (MGS_mean_BS.tif and MGS_std_BS.tif), and</li><li>&nbsp;median grain size (d50_mean_BS.tif and d50_std_BS.tif).&nbsp;</li></ul><p>These grids were the results of the predictive modelling using machine learning models. &nbsp;The inputs to the machine learning models include the sediment samples extracted from the Marine Sediments (MARS) database and a number of environmental and spatial variables. The environmental variables include bathymetry data and its derivatives, data indicating seabed shear stress and data extracted from hydrodynamic modelling. The details on data preparation, modelling and prediction processes can be found in the lineage information.&nbsp;</p><p>The MGS_mean_BS.tif and d50_mean_BS.tif represent the averages of the three MGS grids and the three d50 grids from three different types of machine learning models, respectively. The MGS_std_BS.tif and d50_std_BS.tif represent the standard deviations of the three MGS grids and the three d50 grids from three different types of machine learning models, respectively. The standard deviation grids thus indicate the prediction uncertainties of these two variables.&nbsp;</p><p>These datasets are part of the products delivered by the Resourcing Australia’s Prosperity Initiative (RAPi), under the National resource assessment component and the Offshore Energy Project.</p>