Cartography

Cartography from singlebeam acoustic surveys

During the single beam acoustic field surveys a GPS point is logged every 2 sec, resulting in habitats being attributed from the echogram on average every 2.7 m at a vessel speed of 5 knots. A minimum of six similar attributed points, relating to an average distance of around 15 m, are required for a polygon to be generated. Attributed points are connected at a scale of 1:2,000 to form polygons through a knowledge based interpolation process where the points are joined with similar neighboring transect points to make one habitat polygon.

For shallow Tasmanian coastal waters where rocky reef and unconsolidated habitats are the dominant habitat types this methodology appears to best represent the complex pattern of habitats. In addition, much of the area of the coastal seabed exists within depths that are suitable for defining habitat boundaries with this method with the support of aerial photography (in < 15 m water depth). Greater uncertainties in habitat delineation will exist with single beam acoustic data collection due to the need for interpolation between points.

It is difficult to validate unconsolidated habitats as the boundaries are often not clear on the echogram or visually distinct in the video footage. However, this is generally only an issue in sheltered waters in Tasmania where silt through to sand substrates exist due to the lack of steep depth gradients. This occurs in contrast to the majority of coastal waters that is exposed resulting in few boundaries between unconsolidated substrate, as the substrate is predominantly mapped as sand. Habitat polygons digitised from aerial photographs are displayed in Arc Pad in real time in the field to help identify boundaries in substrate colour and texture. 

Depth measurements taken from the single beam acoustic sounder are used to construct a contour surface. Data are cleaned for erroneous points and are corrected to Mean Sea Level after the depths are adjusted for tidal variation. Arc Info is used to interpolate a TIN (Triangular Irregular Network) surface bathymetric model. Contour lines are generally created for 5 m depth intervals out to the maximum depth of the acoustic transects (usually 40 m).

Cartography from multibeam acoustic surveys

Bathymetric images are generated as a first pass product showing sun-shaded digital elevation models produced using ER Mapper v7.1 based on grids at a spatial resolution of 1 to 3 m depending on the quality of the data. Fledermaus v7.0 software  is then used to generate representative 3D perspective views and profiles across reef systems.

The Simrad EM300 multibeam backscatter data are processed using CMST-GA MB Processv8.11.02.1, a multibeam backscatter processing toolbox co-developed by Geoscience Australia and Centre for Marine Science and Technology (CMST), Curtin University of Technology (Heap et al.,2009). The fully processed backscatter coefficients were corrected for transmission loss and insonification areas based on the equation given in Talukdar et al. (1995). The incidence angle and coordinates on the seafloor, X-Y and depth (Z) were then calculated. The full process within the
toolbox involved the following steps:


1. Conversion from the Simrad raw ALL data format into Matlab data format;
2. Calculation of the absolute X, Y, Z position and the incidence angle θ for each beam and each ping;
3. Removal of the system transmission loss;
4. Removal of the system model;
5. Calculation of the surface backscattering strength, which involves correction for transmission loss and area; and
6. Removal of the angular dependence.


A technique for removing the angular dependence developed by the CMST was applied to the data (cf., Gavrilov et al., 2005). Removing the local mean angular trend also filters out large-scale variations due to change, either sharp or gradual, in the seabed properties along the swath line. To recover this useful information and obtain absolute values of backscatter strength, the angularly equalised backscatter strength within the sampling window is increased by adding the window-mean backscatter level at a specified reference angle (in this case a moderate angle of 25° is used).

The analysis of  backscatter and bathymetric data for characterisation of benthic marine habitats has been the focus of the CERF Marine Biodiversity Hub.


References:

Heap, A.D., Hughes, M., Anderson, T., Nichol, S., Hashimoto, T., Daniell, J., Przeslawski, R., Payne,D., Radke, L., and Shipboard Party, 2009. Seabed Environments and Subsurface Geology of the Capel and Faust basins and Gifford Guyot, Eastern Australia – post survey report. Geoscience Australia, Record 2009/22, 166pp

Gavrilov, A.N., Duncan, A.J., McCauley, R.D., Parnum, I.M., Penrose, J.D., Siwabessy, P.J.W., Woods, A.J., & Tseng, Y-T., 2005. Characterization of the Seafloor in Australia's Coastal Zone using acoustic techniques. Proceedings of the International Conference "Underwater Acoustic Measurements: Technologies & Results", Crete, Greece.