Errors of the digital elevation model

This subchapter discusses a few errors made frequently in the course of digital elevation model construction.

7.4.1. Errors from mistakes made during digitization

It is possible that false height value is associated with the contour line during the vectorizing of the base map.

There can be many reasons for this, some of the - more important - error sources are the following:

• When a new contour line is vectorized the actual layer is not changed - this may occur during the work in AutoCAD.

• False value is written in the data table - frequent during digitization in ArcView.

The digital elevation model created from faulty contours or data tables will be faulty as well therefore the correction of the problematic parts is fundamental. Differences can range from slight to serious faults. In all cases the creation of a “rough” DEM is practical where colour grades have to be given with the same value of the base contour interval of the digitized map because in this way one contour interval will receive one colour and the original contours have to be displayed as well. For example on the map in Figure 8.21 the base contour interval was 25m. When the contour line with the false value is identified it has to be corrected in the data table (for more details regarding identification see chapter 11 of the Geoinformatic Applications textbook).

Correcting the value of the contour lines in ArcView is relatively simple, however, labour consuming. The steps of this are the following (8.2. animation):

1. Prepare the elevation model based on the raw data in the way described above.

2. Display the elevation model and the theme containing the contour lines above it. Adjust the view window so that it occupies the greater part of the screen.

3. Open the data table of the theme containing the contour lines and adjust it so that it occupies the smaller part of the screen and does not cover the view window.

4. Open both for editing (Start Editing).

5. Select the highlight icon (icon 22 in Figure 5.1).

6. Determine the correct value of the wrong contour line (by identifying the surrounding contours).

7. Click on the identified wrong contour line. Data of the selected contour line appears highlighted by yellow colouring in the data table of the contour. Highlighting the table (clicking on its caption) select the edit icon (icon 21 in Figure 7.12) from its icon row then re-write it by clicking into the cell of the value to be corrected in the table.

8. Continue with the identification and correction until all faults are corrected.

9. When completed, close the editing of both the map and the data table (Stop Editing) and save the changes.

10. Delete the elevation model constructed on the basis of faulty values (since the correction of the contour lines does not update the elevation model because they are independent from each other).

11. Create the elevation model - now on the basis of the contour lines containing the corrected values - again.

In Figure 8.21 a trench seems to be visible near the top of a ridge running NW-SE due to a false contour value.

Figure 8.21: Elevation model interpolated from a faulty vector base map

8.2. animation: Correcting the false value of the contour lines

7.4.2. Errors of the interpolation procedure of the software

Significant errors may arise due to the interpolation method of the software. Most frequent problems are the following:

• In the case of an area with irregular shape (e.g. a map segment) data belts impossible to interpret occur in those parts that lack of data. The solution can be the masking out of the unnecessary areas or the cut of the area.

• Even in the case of very differing data density interpolation errors may occur. Most frequent of them appears when less dissected areas are found in-between strongly dissected surfaces. For example when a wide valley crosses a mountainous area. In such cases there are few data in the slightly dissected area (contours are fewer by around a magnitude than in the surrounding area) and as a result false values will occur for these areas. In Figure 8.22 weird stripes are seen in the valleys - marked by red colouring.

• The previous two are combined in the problem of water surfaces: the extended flat surface is not displayed as flat but the relief of the surrounding - dry land - areas is extended over the water surfaces. As a result the water surface will not be flat but almost “waving”.

Figure 8.22: False interpolation in the digital elevation model Controlling questions

Self controlling questions:

How the TIN model is created from the vector file?

How is it possible to construct contour lines from TIN models?

How a GRID model is constructed from a TIN model?

Test:

What does the term TIN mean?

a, Irregular Triangle Network b, Square network

What does the Illuminate Faces option mean at displaying the TIN model?

a, displaying contour lines b, hillshade

In which frame of the TIN model construction window is it possible to define the parameters of the colour fill of the surface?

The basis for our examples is the TIN model. It is important to have the parameters of the map set appropriately (see chapter 11.2 in Geoinformatic Applications) otherwise serious errors may arise. For example, if the height unit is set for metre instead of kilometre the vertical distortion will be 1000-fold as for example 200 metres of rise is counted for 1metre horizontal distance instead of 200 metres rise in the horizontal distance of 1 kilometre!

8.1. 9.1. Slope category map (Slope)

The slope category map shows the slope angle values of the land. Initially select the Surface → Derive Slope…command that opens the window for defining the parameters (Figure 9.1).

Figure 9.1: Setting the parameters of the slope category map

In the Output Grid Specification window (Figure 9.1) the following characteristics can be defined:

• Output GRID Extent: the size of the deduced map can be determined compared to any of the existing layers.

Options are the following: