2004 
  
in be 
scale 
) the 
GPS 
pof 
been 
] use 
aphic 
ns. 
  
tin 
refers 
oto. 
in the 
ed by 
often 
ically 
ity of 
urces, 
only 
id the 
on is 
ch are 
y the 
of the 
X two 
rming, 
lology 
ortho- 
^ field 
age of 
ct that 
ability 
ble to 
Finable 
ew; to 
within 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
cartographic tolerances, only when the aim is to represent 
something which is univocally marked on the territory. Also 
in this sector is nowadays emerging the possibility that some 
feature classes might have overlapping edges whose 
evaluation lies outside statistical methods typical of the 
measurement methodologies, and verges on the incertitude 
peculiar to the fuzzy theories. 
A different evaluation can be made of the existing data set, as 
for example the Corinne LC 1994 and 1999; in such a case is 
a matter of exceptionally complete classifications, capable of 
extremely detailed sub-coding, but produced by means of 
cheaper methodologies and lacking serious controls during 
the implementation phase (manual and automatic photo- 
interpretation derived from low resolution remote-sensing 
images). Specifically the first one, completed in 1994, shows 
an inaccuracy which is far from the tolerance assured by the 
1:100.000 scale and with a percentage of errors and 
omissions such as to make it unsuitable for an environmental 
analysis as detailed as the one proposed in the present study. 
Sample comparisons, carried out in the Park area and in the 
bordering zones have strengthen our decision of proceeding 
to the definition of polygons on an evaluation scale aiming at 
a greater detail and a much lower incertitude. 
GIS IMPLEMENTATION 
The preliminary phase of data input, both geometrical and 
alphanumerical, for the implementation of a G.LS. is 
inevitably a most critical and delicate one, as it constitute the 
metric base of the whole system. In this particular case the 
data deriving from the CAD environment required an 
additional editing and control phase, which sometime proved 
to be complex and lengthy, in order to validate their topology 
and geometrical structure. 
  
  
   
+ % V Se r 
5.9 \ J S oremawo mc) 
  
  
  
Fig.7. Farming land use. 
A graphical entity, depending on the nature and meaning of 
what it represents, may constitute a spatial punctual, linear or 
areal information: the necessity to clearly know the 
functioning of a GIS device is a prerequisite for the 
implementation of subsequent elaborations on a right 
structured Data set. Forests data for example, deriving from 
CAD environment digitalization and including also the 
alphanumerical value, required, as those on land use for 
framing or pasture, a detailed control on possible topological 
inconsistency, in order to be used as information level on 
which to implement the first queries. 
The geometrical analysis verified the closing of single 
entities from an spatial point of view, as the surface of single 
sections needed to be calculated; furthermore a precise 
367 
topological analysis was carried out, with the gradual 
elimination of redundant or incorrect data as for example 
overlapping zones (an area for reforestation cannot be also an 
area for another type of forest), duplicate zones, etc.: such a 
control is necessary in order to guarantee the reliability of 
data (within established tolerances). Before proceeding to the 
connection of the alphanumerical data-base, the correct input 
of such data was verified: if a field is for numeric values 
(area, perimeter, etc.) it cannot be contained in an 
alphanumerical record and vice-versa; if a numeric record as 
a length of n characters, also the other records must be 
consistent with it, etc. When for each class of entities the 
consistency of the data was verified, the alphanumerical data- 
base has been imported in one of the utilized GIS 
environment (Intergraph Geomedia®), and by means of the 
correlation of a common field the geometrical and 
alphanumerical data-bases were linked. Once the starting 
database was completely built, some spatial data were 
automatically calculated (surface, perimeter), and some sub- 
classes were extracted by use of interrogation keys. Both the 
preliminary cleaning operations and the subsequent queries 
were produced in collaboration with the G.LS. experts that 
activated the procedures, working with the same people that 
processed the data, allowing in such a way not only a rapid 
extraction from the G.I.S. of the information required, but 
also a further control on the content and nature of the same 
information: in some cases only after a query of qualitative 
type some gross digitalization mistakes, that could only be 
spotted by a specialist, were identified (e.g. in the forests 
data-warehouse were incorrect 2 on 178x58 fields). 
  
    
   
  
UMCAPPÜLA 
i ^3 
"- | * 
^ SANTERAMOINC/ 
j 7 
  
  
Fig.8. Pasture land use. 
The final phases of the study was characterised by the 
implementation of a Geographical Information System that 
will allow, using suitable visualisation devices (Viewers), the 
complete analysis of the data and information levels 
available through the use of complex query based on the use 
of common logical operators. 
The task of connecting all the basic devices for the 
understanding and monitoring of the territory will rely on the 
GIS elaborations, just as described above, with the ability to 
offer advanced performances in the fields of decision- 
making, planning and territorial management. 
The main feature of the structured System will be: 
- Analysis, also on overlaying levels, of all the acquired 
and univocally referenced cartographic material (current 
vectorial data, historical and current raster data, remote- 
sensed images); 
- Accessibility to all the analyses and studies, with areal, 
linear and punctual reference to the territory, memorised