• Ingen resultater fundet

PROCESSED DATA QUALITY 5.1| BATHYMETRY DATA

In document GEOPHYSICAL SURVEY REPORT (Sider 48-56)

The horizontal and vertical uncertainty of the soundings data were, for the vast majority of the survey area, within the 0.5 m threshold as specified by the client. Sounding density across the routes generally conformed to the revised specification of 16 soundings per 100 cm cell (Figure 20). Density and gap checking was performed on gridded surfaces in Caris HIPS during survey operations for MV Franklin.

Some data gaps exist in the final dataset, these correspond to areas that did not meet the infill threshold criteria (i.e. 4 or more missing 1 m cells that shared a long side) and areas that were flagged as rejected during office data cleaning after both vessels had left the survey area.

Figure 20 Overview of the sounding density surface for Lot 2.

PAGE | 49

Density checks for MV Ping data were performed in the QPS software Qimera using the same 16 soundings per 1 m cell. Data gaps found during survey operations were forwarded to the online survey team from the field office where data processing was performed.

Caris HIPS QC surfaces were generated for all the MV Franklin MBES data within the Lot 2 survey area.

A range of properties are computed for each surface and these are checked systematically to ensure the data falls within specification. The Standard Deviation at 95% confidence interval is checked in order to highlight areas where the vertical spread of soundings within a DTM grid node is high and checks can be made to determine the cause. If necessary, action can be taken to bring the soundings into closer alignment. Regions that have high standard deviations can occur where there are sound velocity errors, errors in the post-processed navigation, where data is acquired in heavy weather and where there are steep slopes such as boulder fields.

Standard deviation surfaces were also checked in Qimera to check the vertical alignment of the MV Ping data.

Figure 21 shows an overview of the Lot 2 Standard Deviation surface for MV Franklin, which presents regions as having low, medium and high standard deviations in green, orange and red, respectively.

Regions where there are numerous boulders are present on all routes but more prevalent on routes 2 and 3. These show up as clusters of orange and red points. These features are not indicative of poor quality data but represent areas that have a greater vertical spread of soundings within a cell relating to the natural roughness of the seabed. Linear artefacts that follow the survey line direction are visible in the standard deviation surface which are the result of minor systematic errors within the data. Such errors relate to small misalignments in vertical position of the vessel and typical differences in sound velocity. An example of such an error can be seen in Figure 22.

Figure 21 Overview of the Standard Deviation surface for Lot 2.

PAGE | 51 Figure 22 Example of good and poor data along Route 3.

Depth convention in Caris HIPS is positive down; vertical exaggeration of cross section is x200. Pink line marks position of cross section.

QC surfaces were computed to show the vertical separation between the mean seabed position and the positions of the shallowest and deepest soundings within a cell. The QC surfaces are used to target both systematic error correction and data cleaning. However, seabed features, such as boulders, are highlighted by these surfaces as well as the outlying soundings. Careful assessment is made of all areas flagged as requiring data cleaning to ensure that real features are not removed from the dataset.

An example of these QC surfaces is shown in Figure 23. Contacts within the boulder area are highlighted in pink and blue since the sounding data deviates from the mean surface by an amount greater than the chosen threshold value for that depth. The surfaces are coloured to indicate the direction of the deviation. Cells where soundings are shallower than the mean surface are highlighted in pink and cells where soundings are deeper than the mean surface are highlighted in blue.

Since the MV Ping survey areas were much smaller than those covered by MV Franklin it was feasible to visually assess the data and use the automatic filters and manual cleaning tools in Qimera to remove outlying soundings.

Figure 23 QC surfaces (pink and blue cells) highlighting boulders along Route 2.

Figure 24 shows the combined vessel TVU surface for Lot 2. The colour scale represents areas where the TVU is low as green, medium as orange, and above the 0.5 m threshold as red. The results show that the TVU values are below the 0.5 m threshold across the full Lot 2 survey area.

An overview of the THU results is shown in Figure 25. The range of values has been restricted to show areas with low THU as blue-green, medium THU as orange and higher THU as red. The overview shows that the values associated with the MV Ping data are relatively higher than those generated by MV Franklin, however all are within the 0.5 m threshold for positional uncertainty. The differences arise as the datasets from MV Franklin and MV Ping are not directly equivalent. Different MBES equipment was used and the data was processed using different software packages (following the same workflow principles). Most importantly though the sizes of the vessels is very different with the smaller vessel, MV Ping, being subjected to larger ranges of motion during the survey of the more dynamic nearshore areas of Route 2 and 5. Faster changes in position resulting from wave motion could lead to elevated uncertainty values for the MV Ping dataset. A close up view of the Route 5 nearshore region is shown in Figure 26. The same pattern is observed in the Route 2 nearshore area.

PAGE | 53

Figure 24 Overview of the Total Vertical Uncertainty (TVU) surface for Lot 2.

PAGE | 55 Figure 26 THU surface at the Nearshore end of Route 5.

In document GEOPHYSICAL SURVEY REPORT (Sider 48-56)