142
C.I.S. Pater
scale, the relative merits of this approach must be considered in relation to the quality
of the original data sources. Sims et al (1995) provides a useful illustration of the
complexity of cross-referencing historical cartographic records.
Serious consideration needs to be given to the relative value of this record to
computer based modelling, with particular reference to the plotting of the gradient
between the intertidal and subtidal areas. With a minimal gradient, contours are
inappropriate in the intertidal area and would not, in any case, be included as a prime
requisite. However, the level of LAT is an important isoline and problems of DTM
construction are encountered if there is minimal data content within the isoline to
provide sufficient z value (node) points for a dense pattern of triangulation, the result
is a ‘flat triangle’ error. Flat triangulation is the process in which a planar triangle
surface is produced by the presence of three identical data values at digitised node
points. For a digitised chart this would occur between isolines of the same depth that
contain no additional spot depths. This does not provide a correct reflection of the
actual surface topography and will affect the value of visualisation and volumetric
analysis, hence the use of linear interpolation to expose these errors which are
particularly acute near the LAT isoline. In order to prevent flat triangulation, the
density of samples must be very high to provide adequate detail of the surface.
However, because of the homogeneous nature of the profile, limited data may be
included on the published chart. In terms of visual analysis of a paper chart, only a
few spot depths or drying heights across a consistent surface in a non-navigational area
provides adequate information.
The provision of data points on charts that can be used as nodes for TIN
generation are therefore insufficient for the generation of viable, functional DTMs.
The only viable DTMs that can be produced are based on collector charts. This
therefore restricts the area of detailed examination from the entire embayment to
locations that have a consistent collector chart survey record, namely the approach
channels to the ports of the Wash. For example, the functional surface produced from
chart 1200 (1993 edition) fails to produce a ‘realistic terrain’ for the approaches to
King’s Lynn (see Area 3 in Fig. 3). The scarcity of soundings included on the chart
for this location, at its worst, represents one data point with a radius of 900 metres.
When the Chart Datum contour is overlain on the DTM (Fig. 6), a clear difference can
be identified between the grey shaded terrain and the white flat areas. Dark spots
represent isolated peaks within intertidal polygons that should have produced
triangulation to form a complete surface both within and without of the intertidal
polygons.
C.I.S. Pater
scale, the relative merits of this approach must be considered in relation to the quality
of the original data sources. Sims et al (1995) provides a useful illustration of the
complexity of cross-referencing historical cartographic records.
Serious consideration needs to be given to the relative value of this record to
computer based modelling, with particular reference to the plotting of the gradient
between the intertidal and subtidal areas. With a minimal gradient, contours are
inappropriate in the intertidal area and would not, in any case, be included as a prime
requisite. However, the level of LAT is an important isoline and problems of DTM
construction are encountered if there is minimal data content within the isoline to
provide sufficient z value (node) points for a dense pattern of triangulation, the result
is a ‘flat triangle’ error. Flat triangulation is the process in which a planar triangle
surface is produced by the presence of three identical data values at digitised node
points. For a digitised chart this would occur between isolines of the same depth that
contain no additional spot depths. This does not provide a correct reflection of the
actual surface topography and will affect the value of visualisation and volumetric
analysis, hence the use of linear interpolation to expose these errors which are
particularly acute near the LAT isoline. In order to prevent flat triangulation, the
density of samples must be very high to provide adequate detail of the surface.
However, because of the homogeneous nature of the profile, limited data may be
included on the published chart. In terms of visual analysis of a paper chart, only a
few spot depths or drying heights across a consistent surface in a non-navigational area
provides adequate information.
The provision of data points on charts that can be used as nodes for TIN
generation are therefore insufficient for the generation of viable, functional DTMs.
The only viable DTMs that can be produced are based on collector charts. This
therefore restricts the area of detailed examination from the entire embayment to
locations that have a consistent collector chart survey record, namely the approach
channels to the ports of the Wash. For example, the functional surface produced from
chart 1200 (1993 edition) fails to produce a ‘realistic terrain’ for the approaches to
King’s Lynn (see Area 3 in Fig. 3). The scarcity of soundings included on the chart
for this location, at its worst, represents one data point with a radius of 900 metres.
When the Chart Datum contour is overlain on the DTM (Fig. 6), a clear difference can
be identified between the grey shaded terrain and the white flat areas. Dark spots
represent isolated peaks within intertidal polygons that should have produced
triangulation to form a complete surface both within and without of the intertidal
polygons.