CIVIL INFOTECH

Remote sensing survey strategy The main strategic issues to consider for the remote sensing (RS) survey are the tools to be used and whether the survey requires full or partial coverage. As indicated in ‘Sampling tools and techniques’ different remote sensing instruments provide different types of data or information, so these must be matched with the information needs identified by the gap analysis. It will be usual to specify which remote sensing technique (or combination of techniques) are to be used in the survey as this can have a major bearing on the RS survey strategy. For aerial and satellite photographic surveys, the default strategy is to achieve full coverage as the swathe width of aerial photography and digital imagery is usually greater than the width of the exposed shore. In topographic and hydrographic LiDAR the default is also for full coverage as the aim is to build a digital terrain model (DTM) showing the topography of the survey area. The survey is conducted in a series of parallel survey lines and any missing lines would leave gaps in the model. The swathe width of the LiDAR instrument can be adjusted to some degree, a wider swath covering more ground at a lower resolution. Hence, if the cost of the survey becomes an issue, the strategic decision would focus on resolution rather than full or partial coverage. Shallow water surveys use a combination of aerial and acoustic remote sensing techniques, and the strategic issues relate to the diminishing efficacy of the aerial techniques as the depth and/or turbidity of the water increases. Acoustic techniques are not compromised by the turbidity of water, so they can be used to complete the survey, if the water depth is sufficient to make their operation logistically feasible. These issues are discussed further in ‘Remote sensing in shallow water’ where two case studies are presented. To complete a DTM, you will need to select aerial and acoustic techniques that supply positional data in three dimensions (X, Y, Z) and ensure they use the same geodetic system (e.g. Latitude, Longitude, WGS84) and that the two data sets are corrected to a common datum point for zero depth before being merged. The issue of full or partial coverage has most significance for swathe acoustic surveys, which are also usually run as a series of parallel lines to allow the build up of a full coverage image. The point is illustrated in the multibeam image from an area of the central English Channel. Fine detail can be mapped over the upper area where full coverage has been achieved, but in the lower area of partial coverage only the larger features that span the data gaps can be mapped. The incentive to move from full to partial coverage is one of survey cost and time.Full coverage is mandatory for finescale mapping and has greatest value in highly heterogeneous areas. It may not be necessary for mapping broader scale features or areas where the substrate is largely homogeneous (e.g. extensive sand plains). Here, quasi ‘full coverage’ can be achieved from less than 100% coverage by interpolating, by eye, across the gaps between survey lines (as seen in the Example of a partial coverage survey). Clearly, this can only be done for features that actually span the gaps, so wider gaps lead to fewer features on the map and more generalised segmentation. It is not usual to interpolate over more than twice the swath width (~33% cover) without supporting inference based on complementary data (usually at a broader spatial scale). Partial surveys over large areas (thousands of square kilometres) may adopt a ‘corridor’ strategy, building full coverage over 1 km wide corridors spaced 5 to 10 km apart. This is a type of nested survey, allowing fine and intermediate scale detail features to be recognised within the corridor and broad scale features to be mapped over the survey area. For more detailed discussion on this topic refer to ‘Partial coverage acoustic surveys’. You should be aware that for hull mounted acoustic sensors, like multibeam and AGDS systems, the swathe width increases with increasing depth of water, while for towed systems that are flown at a constant height above the seabed the swath width is constant (see illustration). Hence, towed systems give greater coverage in shallow water and hull-mounted systems in deeper water. This can influence the choice of acoustic system and the cost effectiveness of the survey. If conditions permit it is recommended to run several acoustic systems simultaneously as they supply complementary data types. It should be left to the survey design stage to determine the spacing to be used between survey lines, But the survey specification should indicate what type of acoustic sensor should be used and whether the survey requires full or partial coverage.