125 
THE USE OF SPATIAL-ORIENTED DATA BASE MANAGEMENT SYSTEMS (DBMS) 
FOR ITALGEO GEOID MANAGEMENT BY GIS 
M. A. Brovelli(*), F. Migliaccio(**) and V. Tornatore(”) 
O Politecnico di Milano, Facoltà di Ingegneria di Como [maria@ipmtf4.topo.polimi.it] 
(**) Politecnico di Milano, DIIAR - Sez. Rilevamento 
[federica@ipmtf4.topo.polimi.it; vin@ipmtf4.topo.polimi.it] 
KEY WORDS: Geographic Information System, Data Base Management Systems, Spatial Extensions, Geoid 
management 
PURPOSE: 
The following work is part of the International Geoid Service (I.Ge.S) project aiming at the use of GIS technology in 
geodetic applications. The variety of the information needed for the calculation and management of the geoid suggests 
the use of a DBMS-oriented storing method instead of a file-based one; this allows for a simpler and best structured way 
to implement the most frequently used actions like updating and querying. The choice of the spatial database (differently 
from other traditional DBMS like, for example, a business administration one) carries the peculiarity of the spatial aspect 
that must be considered while planning the archive itself, in order to obtain a more efficient management system. 
Particularly the relational database management systems supporting the most used GIS-engine interfaces, must have 
extra features such as spatial extensions. In our presentation we will analyse this aspect referring to the construction of a 
geographical information system for the geoidal undulations management. 
1. INTRODUCTION 
In our presentation we will focus on the problems and 
methods related to data structures, e. g. the detailed 
practical description of spatial phenomena, referring to the 
logical level of modelling. So, we are not interested in the 
data modelling of geodetic information, i.e. the set of 
conceptual tools necessary for organising our data. 
To this purpose we would like only to remind that, as in 
territorial information systems devoted to preprocessing 
and analysing other kind of georeferenced data (e. g. 
Land Information Systems -LIS- or Automated Mapping/ 
Facilities Management systems - AM/FM-) the information 
layers we need to describe the data can be vectorial 
(points, arcs or polygons layers) or raster (see Table 1). 
LAYER TYPE 
DATA 
Vector (line) 
Coast lines 
Vector (point) 
Observed gravity anomalies 
Vertical deviations 
Ellipsoidal heights 
(GPS measurements) 
Orthometric heights (levelling 
measurements) 
Digital terrain models 
Batimetry 
Raster 
Gravity anomalies (model) 
Gravity anomalies residual 
terrain corrections 
Geoid estimated by RTC 
Geoid residual calculated via 
Fast Collocation 
Geoid (model) 
ITALGE095 
Digital terrain models 
Crustal density models 
Moho depth models 
Tab. 1 - Examples of vector and raster layers. 
If we look, for instance, at the coastal lines and we 
consider the subdivision between land and sea, we may 
introduce a vectorial complex polygons layer for the 
geometrical and topological description of the territory. 
By the way, the majority of our data may be modelled by 
means of point entities or cell layers. To the first group 
belong observations of different kinds (for instance the 
gravity anomalies or the geoid heights obtained by 
comparison of GPS measurements and levelling) and 
data interpolated in sparse points. To the second group 
belong all gridded data, such as the reference geoid 
computed by global collocation (e.g. the ITALGE095 
geoid), the DTMs and the map of crustal density. So, with 
respect to data manipulation, extraction and recovering 
we have to face all the problems related with the 
geographical features description and we need to 
interface the GIS engine we have chosen to adopt for 
geodetic purposes with an RDBMS which presents 
special extensions allowing for an efficient management 
of our database. 
2. THE DBMS APPROACH 
As referred in previous works (Biagi et al, 1998; Biagi et 
al, 1999) the GIS we have selected in order to customize 
it for geodetic purposes is the GRASS system. 
This system was originally designed and developed by the 
Environmental Division of the U. S. Army Construction 
Engineering Research Laboratory. Since the first release 
in 1985, it was distributed with source code, in such a way 
that users from different sides (education and research 
institutions, government organizations and private firms) 
and with various interests (land planners, landscape 
architects, geographers, geologists, ...) may customize 
this software to settle their own questions. Furthermore, 
as it was (and is) freely available, the number of users 
and the types of applications in the research environment 
have rapidly grown. From 1997 the GRASS development, 
support and enhancement of public domain version, was 
undertaken by the GRASS Research Group - Department