Academia.eduAcademia.edu

Figure 9 – uploaded by Charitha Pattiaratchi

See full PDFdownloadDownload figure
Fig. 10. Probability density distribution of the mean BS intensity values associated with the underwater geomorphic features mapped on the MBES bathymetry models. Whilst rocky substrates show the highest range of mean relative BS strength, sandy substrates show lower values over a wider range which is representative of a variety of sediment grain sizes and seabed roughness. The remaining geomorphic features mapped at Geraldton (sand bars and sheets, nearshore zone, rippled sand flats and sand dunes) are Areas mapped as shallow limestone reef and low relief substrate show the highest range of mean relative BS strength (Fig. 10) and these rocky areas often host ALGAE and SEAGRASS. SEAGRASS and ALGAE recorded the highest relative BS strength. It is not evident whether the nature of the substrate or the actual benthic cover is most influential on the acoustic properties, but they seem to determine together the seabed acoustic response.

Figure 10 Probability density distribution of the mean BS intensity values associated with the underwater geomorphic features mapped on the MBES bathymetry models. Whilst rocky substrates show the highest range of mean relative BS strength, sandy substrates show lower values over a wider range which is representative of a variety of sediment grain sizes and seabed roughness. The remaining geomorphic features mapped at Geraldton (sand bars and sheets, nearshore zone, rippled sand flats and sand dunes) are Areas mapped as shallow limestone reef and low relief substrate show the highest range of mean relative BS strength (Fig. 10) and these rocky areas often host ALGAE and SEAGRASS. SEAGRASS and ALGAE recorded the highest relative BS strength. It is not evident whether the nature of the substrate or the actual benthic cover is most influential on the acoustic properties, but they seem to determine together the seabed acoustic response.

Related Figures (16)

Fig. 1. Map of the study area, showing extent of the multibeam survey and locations of sediment samples and high resolution video transects.
Fig. 1. Map of the study area, showing extent of the multibeam survey and locations of sediment samples and high resolution video transects.
Fig. 2. W-E profile of the Geraldton coastal platform derived from the multibeam bathymetry. Sand bars and sheets (Fig. 3a and b): are sand bodies showing higher elevation compared to the surrounding substrates due to sediment ac- cumulation. The difference in elevation to the surrounding seabottom is up to 1 m for sand bars, but lower thicknesses were found for sand The MBES data collected through the QUINSy software interface vere filtered in real time for signal spikes through a manually selected icquisition range based on seafloor bathymetry. The dynamic correc- ion for boat pitch, roll and heave was also applied in real time in the ield. A further refinement of the data to remove spikes’ was carried ut at the Department of Transport processing centre in Fremantle, ising the QLOUD software interface. Digital Elevation Models (DEMs) )f MBES bathymetry were imported into a GIS project specifically devel- yped for this data and polygons were manually digitised on the DEMs at | fixed scale of 1:2000 which was found to be the most convenient vorking scale considering data resolution and morphological feature di- nensions. During the mapping operations, the MBES BS data were also ‘ompared to the bathymetry to correlate the two datasets.
Fig. 2. W-E profile of the Geraldton coastal platform derived from the multibeam bathymetry. Sand bars and sheets (Fig. 3a and b): are sand bodies showing higher elevation compared to the surrounding substrates due to sediment ac- cumulation. The difference in elevation to the surrounding seabottom is up to 1 m for sand bars, but lower thicknesses were found for sand The MBES data collected through the QUINSy software interface vere filtered in real time for signal spikes through a manually selected icquisition range based on seafloor bathymetry. The dynamic correc- ion for boat pitch, roll and heave was also applied in real time in the ield. A further refinement of the data to remove spikes’ was carried ut at the Department of Transport processing centre in Fremantle, ising the QLOUD software interface. Digital Elevation Models (DEMs) )f MBES bathymetry were imported into a GIS project specifically devel- yped for this data and polygons were manually digitised on the DEMs at | fixed scale of 1:2000 which was found to be the most convenient vorking scale considering data resolution and morphological feature di- nensions. During the mapping operations, the MBES BS data were also ‘ompared to the bathymetry to correlate the two datasets.
Fig. 3. Examples of the underwater morphology features mapped at Geraldton on the high resolution DEMS of multibeam bathymetry. A) Sand sheet adjacent to an area of low relief substrate; B) Sand bars with rippled sand in the bar's trough; C) Nearshore beach zone adjacent to a rippled sandy area which evolves into underwater sand dunes further offshore; D) Shallow limestone reef system surrounded by an area of low relief substrate. These features were described in Sections 4.1.1, 4.1.2 and 4.1.3.
Fig. 3. Examples of the underwater morphology features mapped at Geraldton on the high resolution DEMS of multibeam bathymetry. A) Sand sheet adjacent to an area of low relief substrate; B) Sand bars with rippled sand in the bar's trough; C) Nearshore beach zone adjacent to a rippled sandy area which evolves into underwater sand dunes further offshore; D) Shallow limestone reef system surrounded by an area of low relief substrate. These features were described in Sections 4.1.1, 4.1.2 and 4.1.3.
Fig. 4. Underwater morphology mapping at Geraldton based on DEMs of multibeam bathymetry. This figure shows part of the analysed areas to provide an example of the mapping outcomes. The blue tones of the colouring scale indicate the artificial elements (i.e. aquaculture zone, dredged channel, dredged port basin, and wrecks); yellow to red colours indicate uncolonised sandy substrates (i.e. nearshore zone, rippled sand flat, sand bars and sheets, and underwater sand dunes); the shallow limestone reefs and low relief substrate areas are coloured green and grey respectively.
Fig. 4. Underwater morphology mapping at Geraldton based on DEMs of multibeam bathymetry. This figure shows part of the analysed areas to provide an example of the mapping outcomes. The blue tones of the colouring scale indicate the artificial elements (i.e. aquaculture zone, dredged channel, dredged port basin, and wrecks); yellow to red colours indicate uncolonised sandy substrates (i.e. nearshore zone, rippled sand flat, sand bars and sheets, and underwater sand dunes); the shallow limestone reefs and low relief substrate areas are coloured green and grey respectively.
Summary of the main characteristics of the Geraldton sediment facies. Table 1
Summary of the main characteristics of the Geraldton sediment facies. Table 1
Fig. 5. Microscope photos of loose grains of the following sediment facies: A) Fine modern skeletal sand; B) Medium-coarse modern and relict skeletal sand; C) Coarse quartzose and relict skeletal sand; D) Cross-shore transect showing the distribution of sediment facies with depth; E) Line plots of sediment grain size across depth showing percentage of fine, medium and coarse sands; F) Line plots across depth showing carbonate content (%CaCO3), percentage of modern skeletal grains and percentage of relict carbonate grains (intraclasts + relict skeletal grains). Whilst this profile represents an average sediment distribution across depth throughout the study area, off the Chapman River mouth the “fine-medium quartzose and modern skeletal sand” sediment facies dominates the coastal platform.
Fig. 5. Microscope photos of loose grains of the following sediment facies: A) Fine modern skeletal sand; B) Medium-coarse modern and relict skeletal sand; C) Coarse quartzose and relict skeletal sand; D) Cross-shore transect showing the distribution of sediment facies with depth; E) Line plots of sediment grain size across depth showing percentage of fine, medium and coarse sands; F) Line plots across depth showing carbonate content (%CaCO3), percentage of modern skeletal grains and percentage of relict carbonate grains (intraclasts + relict skeletal grains). Whilst this profile represents an average sediment distribution across depth throughout the study area, off the Chapman River mouth the “fine-medium quartzose and modern skeletal sand” sediment facies dominates the coastal platform.
Fig. 6. Fossil composition of the modern skeletal grain sediment fractions for A) Fine modern skeletal sand and B) Medium-coarse modern and relict skeletal sand
Fig. 6. Fossil composition of the modern skeletal grain sediment fractions for A) Fine modern skeletal sand and B) Medium-coarse modern and relict skeletal sand
Fig. 7. Photos of representative habitat types used for acoustic habitat mapping.
Fig. 7. Photos of representative habitat types used for acoustic habitat mapping.
Fig. 8. Angular response (backscatter versus incidence angle) curves of the main seabed types found in the study area. Seagrass meadows (SEAGRASS), macroalgal communities (ALGAE) and mixed low density macroalgal and seagrass communities (MIXED) show similar curves, but the sand substrates (SAND) curve is remarkably different. a, > aa ea a, <a The shape of mean angular response curves for SEAGRASS, ALGAE and MIXED habitats produced a similar profile (Fig. 8), with SEAGRASS showing some 0.5 dB higher relative BS strength than ALGAE. Fig. 9 shows the probability density distribution of the mean relative BS strength associated with: ALGAE, SEAGRASS, SAND and MIXED, and it confirms the overlap between the relative BS strength of SEAGRASS and ALGAE, indicating that the identified biota is not individually driving the acoustic response of the seabed. These habitats occur over acoustically similar regions of the seabed, mostly hardgrounds with shallow consoli- dated sandy cover. The slight increment in relative BS strength linked to seagrass cover might be associated with the presence of gas bubbles and dense root structure as previously speculated by Parnum (2007), Wilson and Dunton (2009) and De Falco et al. (2010). SAND record the lowest relative BS strength measured and a significantly different shape of the angular response curve. There is general agreement that harder and rougher substrates such as hardgrounds produce stronger backscat- ter signals than softer and smoother substrates. MIXED has intermediate relative BS strength between SAND and ALGAE, and the shape of the angular response curve of MIXED is also similar to the ALGAE and SEAGRASS curves. The probability density distribution of the mean rela- tive BS strength measured for this habitat shows a significant overlap with SEAGRASS and ALGAE (Fig. 9) indicating that the density of the mapped biota is not the only factor influencing the acoustic response of these communities. MIXED is associated with a variety of substrates rang- ing from hardgrounds with highly consolidated sandy cover to low thick- ness sand sheets characterised by different grain sizes. This may explain the wide range of the probability density distribution curve of mean rela- tive BS strength.
Fig. 8. Angular response (backscatter versus incidence angle) curves of the main seabed types found in the study area. Seagrass meadows (SEAGRASS), macroalgal communities (ALGAE) and mixed low density macroalgal and seagrass communities (MIXED) show similar curves, but the sand substrates (SAND) curve is remarkably different. a, > aa ea a, <a The shape of mean angular response curves for SEAGRASS, ALGAE and MIXED habitats produced a similar profile (Fig. 8), with SEAGRASS showing some 0.5 dB higher relative BS strength than ALGAE. Fig. 9 shows the probability density distribution of the mean relative BS strength associated with: ALGAE, SEAGRASS, SAND and MIXED, and it confirms the overlap between the relative BS strength of SEAGRASS and ALGAE, indicating that the identified biota is not individually driving the acoustic response of the seabed. These habitats occur over acoustically similar regions of the seabed, mostly hardgrounds with shallow consoli- dated sandy cover. The slight increment in relative BS strength linked to seagrass cover might be associated with the presence of gas bubbles and dense root structure as previously speculated by Parnum (2007), Wilson and Dunton (2009) and De Falco et al. (2010). SAND record the lowest relative BS strength measured and a significantly different shape of the angular response curve. There is general agreement that harder and rougher substrates such as hardgrounds produce stronger backscat- ter signals than softer and smoother substrates. MIXED has intermediate relative BS strength between SAND and ALGAE, and the shape of the angular response curve of MIXED is also similar to the ALGAE and SEAGRASS curves. The probability density distribution of the mean rela- tive BS strength measured for this habitat shows a significant overlap with SEAGRASS and ALGAE (Fig. 9) indicating that the density of the mapped biota is not the only factor influencing the acoustic response of these communities. MIXED is associated with a variety of substrates rang- ing from hardgrounds with highly consolidated sandy cover to low thick- ness sand sheets characterised by different grain sizes. This may explain the wide range of the probability density distribution curve of mean rela- tive BS strength.
Fig. 9. Probability density distribution of the mean relative BS strength values associated with the main habitat categories investigated in this study. This figure outlines the similar- ity of the acoustic response of seagrass meadows (SEAGRASS), macroalgal communities (ALGAE) and mixed low density macroalgal and seagrass communities (MIXED).
Fig. 9. Probability density distribution of the mean relative BS strength values associated with the main habitat categories investigated in this study. This figure outlines the similar- ity of the acoustic response of seagrass meadows (SEAGRASS), macroalgal communities (ALGAE) and mixed low density macroalgal and seagrass communities (MIXED).
Fig. 11. Benthic habitat map based on the results of the supervised classification of a raster of angle-independent BS levels. Seagrass percentage of cover and mean benthic cover based on the analysis of video transects are overlayed on the classified acoustic map. Due to the similarity of the acoustic response from SEAGRASS and ALGAE, the mapping of these habitats could not be performed separate- ly by applying classification algorithms to a raster grid derived from angle-independent backscatter levels only but these communities were mapped together as “dense vegetation” and MIXED is called “sparse vegetation” on the map legend for simplicity. Seagrass and macroalgae percentage of cover determined from the analysis of quan- titative video transects were also overlayed on the classified acoustic map, allowing the distinction between dense seagrass or macroalga communities to be visualised (Fig. 11). The benthic habitat map produced from the supervised classification of a raster of angle- independent BS levels (Fig. 11) only achieved an overall accuracy of indicative of sandy substrates, and record a much lower relative BS strength over a wider range of values (Fig. 10). The sand bars and sheet features have shown the lowest standard deviation of the range of BS intensity measured compared to the remaining sandy features with an average BS intensity value of approximately — 16 dB. This might be due to the lower sediment thickness of the areas mapped as nearshore zone, rippled sand flats and sand dunes, which can be ascertained from the observation of bathymetric data. Of note is the bi- modal distribution of the relative BS strength extracted from the areas mapped as rippled sand flats. There are two possible causes for this re- sult, as these areas could be characterised by two different roughness or grain size classes which would result in two modes in the relative BS strength measurements.
Fig. 11. Benthic habitat map based on the results of the supervised classification of a raster of angle-independent BS levels. Seagrass percentage of cover and mean benthic cover based on the analysis of video transects are overlayed on the classified acoustic map. Due to the similarity of the acoustic response from SEAGRASS and ALGAE, the mapping of these habitats could not be performed separate- ly by applying classification algorithms to a raster grid derived from angle-independent backscatter levels only but these communities were mapped together as “dense vegetation” and MIXED is called “sparse vegetation” on the map legend for simplicity. Seagrass and macroalgae percentage of cover determined from the analysis of quan- titative video transects were also overlayed on the classified acoustic map, allowing the distinction between dense seagrass or macroalga communities to be visualised (Fig. 11). The benthic habitat map produced from the supervised classification of a raster of angle- independent BS levels (Fig. 11) only achieved an overall accuracy of indicative of sandy substrates, and record a much lower relative BS strength over a wider range of values (Fig. 10). The sand bars and sheet features have shown the lowest standard deviation of the range of BS intensity measured compared to the remaining sandy features with an average BS intensity value of approximately — 16 dB. This might be due to the lower sediment thickness of the areas mapped as nearshore zone, rippled sand flats and sand dunes, which can be ascertained from the observation of bathymetric data. Of note is the bi- modal distribution of the relative BS strength extracted from the areas mapped as rippled sand flats. There are two possible causes for this re- sult, as these areas could be characterised by two different roughness or grain size classes which would result in two modes in the relative BS strength measurements.
Error matrix derived from a reinterpretation of Table 2, showing the mapping accuracy increase which could be achieved by restricting the habitat categories to sand and vegeta- tion only.
Error matrix derived from a reinterpretation of Table 2, showing the mapping accuracy increase which could be achieved by restricting the habitat categories to sand and vegeta- tion only.
Error matrix for the benthic habitat map resulting from the supervised classification of an- gle-independent BS levels shown in Fig. 11. The bottom line summarises the classification accuracy for each habitat. ~54% (Table 2), with the highest classification accuracy recorded for SAND (~79%), followed by sparse vegetation (~52%) and dense vegeta- tion (~25%). As shown in Table 3, a higher mapping accuracy of 78% could be achieved by limiting the habitat categories to sand and vegeta- tion only. However the overall study benefits from having information on the density of the main biota on the benthic habitat map. The in- creased mapping accuracy attainable by mapping sand and vegetation only indicates that a raster of angle-independent BS levels does not con- tain sufficient information to discriminate between different densities of seagrass or macroalgal communities.
Error matrix for the benthic habitat map resulting from the supervised classification of an- gle-independent BS levels shown in Fig. 11. The bottom line summarises the classification accuracy for each habitat. ~54% (Table 2), with the highest classification accuracy recorded for SAND (~79%), followed by sparse vegetation (~52%) and dense vegeta- tion (~25%). As shown in Table 3, a higher mapping accuracy of 78% could be achieved by limiting the habitat categories to sand and vegeta- tion only. However the overall study benefits from having information on the density of the main biota on the benthic habitat map. The in- creased mapping accuracy attainable by mapping sand and vegetation only indicates that a raster of angle-independent BS levels does not con- tain sufficient information to discriminate between different densities of seagrass or macroalgal communities.
Fig. 12. Seabed mobility map of the Geraldton coastal system. The identified geomorphic features provide useful insight in regard to sediment dynamics. Areas of sediment deposition at indicated by the sand bars and sheet system. Stable areas in terms of sediment transport are limestone reefs and flat or low sloping areas with a shallow consolidated sandy cover. Ripple sand flats and underwater sand dunes indicate sediment transport by bottom currents, and consequently are indicated as areas where potential sediment transfer may occur.
Fig. 12. Seabed mobility map of the Geraldton coastal system. The identified geomorphic features provide useful insight in regard to sediment dynamics. Areas of sediment deposition at indicated by the sand bars and sheet system. Stable areas in terms of sediment transport are limestone reefs and flat or low sloping areas with a shallow consolidated sandy cover. Ripple sand flats and underwater sand dunes indicate sediment transport by bottom currents, and consequently are indicated as areas where potential sediment transfer may occur.
Fig. 13. Sketch illustrating the nature of shallow water carbonate sediment production in the studied shallow water warm-temperate marine embayments. Both the macroalgal and seagrass carbonate factories are producing similar, but somewhat different, carbonate sediments with higher production rates for seagrass meadows. A lack of mixing of the particles de- riving from these benthic communities occurs in the study area, with fine sands dominating the seagrass sites and medium-coarse sands occurring where macroalgal communities are prevalent. The diagram also summarises the substrate related partitioning of these communities with macroalgae located on the high energy shallow limestone reef systems at the sea- ward edge of the coastal platform and seagrasses rooted in the bedrock depressions of flat or low sloping hardgrounds at sheltered locations close to shore.
Fig. 13. Sketch illustrating the nature of shallow water carbonate sediment production in the studied shallow water warm-temperate marine embayments. Both the macroalgal and seagrass carbonate factories are producing similar, but somewhat different, carbonate sediments with higher production rates for seagrass meadows. A lack of mixing of the particles de- riving from these benthic communities occurs in the study area, with fine sands dominating the seagrass sites and medium-coarse sands occurring where macroalgal communities are prevalent. The diagram also summarises the substrate related partitioning of these communities with macroalgae located on the high energy shallow limestone reef systems at the sea- ward edge of the coastal platform and seagrasses rooted in the bedrock depressions of flat or low sloping hardgrounds at sheltered locations close to shore.
Related topics:
Earth SciencesMarine GeologyWave EnergyFishery ManagementCoastal and estuarine ProcessesWave Exposure
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved