Data from multibeam echosounder surveys taken as part of the Ningaloo Outlook project are classified into various seafloor cover types according to their hardness, rugosity and depth. The classifications are validated with towed video ground truth where it is available. Three AOIs are classified, two that were explicitly part of the Ningaloo Outlook Deep Reefs project and a third transect that was acquired incidentally while RV Investigator was transiting between locations. Due to the nature of the acquired data, two different approaches were taken for the classification, the first approach used multibeam backscatter angular response curves along with rugosity as input to a maximum likelihood classifier. The second approach used flattened multibeam backscatter (i.e. with the angular effects removed), along with rugosity as inputs to a Random Forest Classifier. Estimates of the accuracy of the classifiers are produced, where possible, along with area statistics for the different substratum observed in the classified maps. The maps are GeoTIFFs with text based classification keys. [Note:] Area 3a only

Data from multibeam echosounder surveys taken as part of the Ningaloo Outlook project are classified into various seafloor cover types according to their hardness, rugosity and depth. The classifications are validated with towed video ground truth where it is available. Three AOIs are classified, two that were explicitly part of the Ningaloo Outlook Deep Reefs project and a third transect that was acquired incidentally while RV Investigator was transiting between locations. Due to the nature of the acquired data, two different approaches were taken for the classification, the first approach used multibeam backscatter angular response curves along with rugosity as input to a maximum likelihood classifier. The second approach used flattened multibeam backscatter (i.e. with the angular effects removed), along with rugosity as inputs to a Random Forest Classifier. Estimates of the accuracy of the classifiers are produced, where possible, along with area statistics for the different substratum observed in the classified maps. The maps are GeoTIFFs with text based classification keys. [Note:] 120m Transect only

Data from multibeam echosounder surveys taken as part of the Ningaloo Outlook project are classified into various seafloor cover types according to their hardness, rugosity and depth. The classifications are validated with towed video ground truth where it is available. Three AOIs are classified, two that were explicitly part of the Ningaloo Outlook Deep Reefs project and a third transect that was acquired incidentally while RV Investigator was transiting between locations. Due to the nature of the acquired data, two different approaches were taken for the classification, the first approach used multibeam backscatter angular response curves along with rugosity as input to a maximum likelihood classifier. The second approach used flattened multibeam backscatter (i.e. with the angular effects removed), along with rugosity as inputs to a Random Forest Classifier. Estimates of the accuracy of the classifiers are produced, where possible, along with area statistics for the different substratum observed in the classified maps. The maps are GeoTIFFs with text based classification keys. [Note:] Area 5 only

The remote Kimberley coast of north-western Australia is one of the few marine environments domains on earth largely unaffected by human use. However, the region is undergoing increasing economic importance as a destination for tourism and significant coastal developments associated with oil and gas exploration. The objective of the project was to reconstruct a timeline of inferred water quality changes from the sediment record for a selected set of sites in the Kimberley, Western Australia. The project made use of palaeoecological approaches to reconstruct a chronology of change over the last approximately 100 years using a series of biogeochemical proxies for phytoplankton composition and biomass, temperature and terrestrial influences. Where possible these were matched to historical land/water use, meteorological or hydrological observational records. The project examined sediment cores from three coastal locations, Koolama Bay (King George River), Cygnet Bay and Roebuck Bay. Each sampling location provided a contrast with which to evaluate changes over either a spatial or temporal gradient of human or natural influence. Sediment cores (up to 1.5 m) were obtained from each of these locations in the expectation that they would provide a time series for about the last 100 years. A set of parameters was measured along the core length (every 1-2 cm) for some or all cores depending on the particular focus for the location: 210Pb and 137Cs; 15N isotope; 13C isotope; Carbon/Nitrogen ratio; Sedimentation rate and grain size; Total Organic Carbon (TOC) and Total Nitrogen (TN); Biosilicate; Biomarkers; TEX86; long chain n-alkanes (C27+C29+C31); Elemental carbon (or black carbon). Rainfall data was obtained from the Australian Bureau of Meteorology website (www.bom.gov.au). Stream flow data was obtained from the Western Australian Department of Water website (www.water.wa.gov.au). Historical bushfire data was obtained from the Western Australian Department of Parks and Wildlife. The metadata record only relates to data generated as part of the sediment analysis.