originally captured. 
Random 
The points were distributed randomly. Data 
acquisition technique such as field surveying (total 
station), photogrammetry (random sampling) 
produces this kind of pattern. The problem with this 
pattern is that the topology of the points is not easy 
to defined. 
Contour 
This is another common way of storing DTM data. 
Points with equal height were depicted in one same 
line, ie. known as a contour line. Contours are 
usually depicted on topographic maps. The contours 
normally converted to digital form by manual or 
automatic digitizing for further processings. Digital 
contours data is easy to handle and manipulate, for 
instance in DTM. Its easiness and simplicity to 
handle, made the data popular for DTM generation. 
Triangular Irregular Network (TIN) 
By definition a TIN is formed by a series of irregular 
triangles. Each triangle consists of three nodes, and 
each node has x, y, z coordinates. These nodes 
represent data points, and they are not interpolated 
(i.e. every point is utilized). For example, the nodes 
may come from measured random spot heights of 
field surveying, or they may come from selective 
photogrammetric sampling, or they may be the result 
of digitizing of contours, etc. A connection of any 
three ‘proper’ nodes forms a triangle, and if further 
connections are made to these data points, then a 
network called TIN is established. The word 
‘proper’ in the preceding line, means a process of 
connecting the three nodes by a systematic 
triangulation algorithm, e.g. using Delaunay 
triangulation (Huber, 1995). The TIN data structure 
actually has been used for terrain representation for 
quite sometime, back to 1970's (Peucker et al, 1978). 
Now the structure is being utilized in quite a number 
of GIS and mapping software (Kumler, 1994). The 
structure has the following advantages: 
* the size and shapes of the triangles varies 
according to terrain roughness (i.e. in flat 
surface, the size tend to be larger than in 
rougher terrain). 
* ability to handle breaklines and specific 
points into the triangle network. 
* ‘saddle point problem’ as in grid structure 
can be avoided. 
8 
* some of the computation tasks are easy to 
perform, e.g. slope and aspect calculation. 
  
  
ee 
Contours 
  
  
  
Figure 1 Some of the most prominent DTM data 
structures. 
Besides the advantages offered by the TIN, it has 
some limitations, e.g. need more computer storage 
space to store the triangle nodes, triangle sides, and 
the topology information. In this aspect, the grid is 
much simpler and takes less space. 
Since the structure (ie. TIN) offers several 
advantages, thus we adopted it in an attempt to 
identify the most developable land. 
TIN-based Applications 
This section reviews some of the typical DTM 
applications, namely, contouring, and. slope-aspect 
mapping. 
Contouring. One of the common output of DTM is 
contour map. The contours may be generated in 
different ways, but in TIN it is appropriate to 
interpolate the contours using simple method (e.g. 
linear interpolation) then followed by line smoothing 
(Alias, 1992). The contour interval also play a role. 
Larger interval may represent course terrain, on the 
other hand, the smaller interval may effect the 
storage, interpolation speed, and producing touching 
contour lines in steep slopes. So the chosen interval 
should reflect what we are going to do with the 
contours. Generally, application which needs 
detailed information should go for smaller interval, 
whereas larger interval for the less detail. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996 
Slope: 
eleme: 
charac 
standa 
slope 
forwat 
nodes, 
Then ! 
directi: 
The sti 
north 1 
undul 
appro 
topogt 
Conto! 
manue 
the ma 
such a 
reserve 
also ex 
availal 
  
Fi 
The g 
develo 
area). 
carriec 
overla 
operat 
The fo 
land ic