Appli- 
quares 
on, In- 
temote 
ompu- 
'y, Na- 
ry, El- 
stellte 
n Ob- 
on, Fa- 
Stutt- 
Recon- 
Jiscon- 
Intern. 
e Sen- 
than. 
nd In- 
Robust 
(Eds.), 
INTEGRATION OF DTM DATA STRUCTURES INTO GIS DATA MODELS 
D. Fritsch and A. Pfannenstein 
Chair for Photogrammetry and Remote Sensing 
Technical University Munich 
Arcisstr. 21, D-8000 Munich 2, Germany 
Tel: + 49-89-2105 2671; Fax: + 49-89-280 95 73; Telex: 522854 tumue d 
E-mail: anton@photo.verm.tu-muenchen.de 
Commission III 
ABSTRACT: 
Most of the Geographical Information Systems (GIS) 
used in practice have been designed for management 
and analyses of planimetric data. In the most recent 
2D * 1D approach the 1D is given by a digital terrain 
model (DTM) more or less isolated from the planimetric 
data (2D). On the contrary there is an increasing de- 
mand for 3D data models and analyses. This requires the 
unification of planimetric and height data management 
within one GIS data model. 
A possible way is the integration of DTM data structures 
into GIS data models which is treated by this paper. It 
starts with an overview of DTM data structures and 
DTM data storage. With regard to GIS data models 
concepts for the efficient integration of DTM data struc- 
tures into GIS are given. The paper concludes with first 
assessments considering system response, data consi- 
stencies and data analyses. 
Key words: digital terrain model, geographical informa- 
tion system, data models, data analyses. 
1. INTRODUCTION 
Since more than 30 years digital terrain modelling is 
under research not only at photogrammetric institutions 
but also in civil engineering and in industry. Digital 
terrain models (DTM) represent the earth surface by 
boundary descriptions in different data structures. The 
terrain heights z are variables of fixed planimetric data 
Xy, thus the three-dimensional problem is shrinked to a 
one-dimensional one. Corresponding data models are 
very simple: topology is maintained by grid and triangle 
structures, and object semantics characterizes only the 
nodes within terrain. 
497 
When looking into the history of GIS these systems 
evolved mainly from classical two-dimensional mapping 
problems (Bill/Fritsch,1991). The reference surface is 
given by the projection surface of the underlying coordi- 
nate system, onto which the three-dimensional data is 
projected. Therefore, they are restricted in geometry on 
two variable dimensions (e.g. x,y). Height data, if at all, 
are supplementary information of the situation data. 
But the earth is not flat. So, GIS and DTM should not 
longer be seen isolated. One way out of the dilemma is 
to interface DTM program packages with GIS which is 
described in Ebner et al. (1990), Fritsch (1990b,1991), 
and Reinhardt (1991). However, the use of two separa- 
ted databases rises the problem of data consistency, 
which can only be solved if data structures for both, 
DTM and GIS are integrated in one data model. 
The integration of a DTM into a GIS demands for a total 
unification of the underlying data sets as well as the 
methods being applied. Two main procedures can be 
used which may also be combined with each other 
(Fritsch,1990a) 
- three-dimensional coordinates for all geographic 
elements 
- digitalterrain models as constituents of a geographic 
database 
While the first approach is costly in terms of storage 
elements - the situation elements have to be supplemen- 
ted by means of additional height elements - the latter 
one is easier in concept and realization. Furthermore, 
situation data are dense only in densely populated re- 
gions, however, in agricultural areas as well as poor 
populated regions coarse distribution of planimetry 
must be overcomed. Therefore the integration of a DTM 
into a GIS is also more pragmatic from this point of view. 
The following constraints have to be considered: On the 
one hand data storage should be not redundant leading