565 
of establishing 
the user. This 
ata included and 
11 determine the 
ation generated, 
red in terms of 
racy is relative 
g. 3). 
ergy invested in 
ining, entering, 
ul data exist in 
aphical, digital 
1. A primary 
ore, becomes the 
ns into a single 
ayers, fields or 
spatial entities 
ristics of the 
be utilized in 
tual variable or 
tr value, and 2) 
it resides in 
). 
ition in one of 
5, surfaces, and 
leir geographic 
isic techniques: 
Ld cell format. 
x,y coordinates 
s or polygon. 
:al format are 
'or line (Marble 
located by its 
lattice or grid 
mation or geo- 
md limitations, 
iformation more 
need data in N 
g. x,y) and is 
nalysis (Marble 
results in less 
more computer 
format results 
city, however, 
ncy is higher. 
points whose 
er of axis in 
represents a 
t position in 
. Spatial data 
ted with raster 
istration, yet, 
Environmental 
Satellite Image 
Data 
nage processing 
3 
is positionally less accurate. The strengths of 
both formats (grid cell and x,y coordinates) can be 
utilized if the GIS is capable of dealing with both 
types of data and providing a means of conversion 
from one structure to the other (Myers, 1983, 1985; 
Tomlinson, 1983). In an applied situation, such as 
AISC, vector format is recommended for data storage 
and some data array manipulations. These include 
multivariate analysis of precipitation, temperature, 
soil moisture, terrain and ground cover data. The 
raster format includes AVHRR type data for overlay 
and display. For spatial analysis, vector data 
formats can be readily converted to raster in 
modeling procedures. 
7 DATA BASE MANAGEMENT 
A GIS must have the capabilities of supporting 
numerous data bases that are equally accessible by 
multiple users. An effective data base management 
system enables efficient data storage, retrieval and 
update (fig. 5). 
Figure 5 
7.1 Data Accessibility and Telecommunication 
Pertinent data collection and processing 
capabilities may not always reside within the 
particular agency interested in performing a 
specific task. In reality, information and specific 
processing capabilities are located at various 
national and international institutions. Further, 
not all institutions incorporate similar standards 
in terms of data format, data information content, 
or computing hardware and software. Telecommunica 
tion links and associated communication software 
enables data format transformation, transfer and 
processing to take place thus strengthening the 
capabilities of the single agency. The concept of 
telecommunication linkage brings together 
information and data processing capabilities needed 
in complex problem solving (fig. 6). 
7.2 Data Analysis 
Data analysis procedures enable the GIS user to 
statistically manipulate data sets so the specific 
information can be generated for particular 
interpretation. The procedures range from simple 
overlaying of data sets to complex statistical 
functions integrated into sophisticated spatial 
models. For example, canonical analysis can be 
applied to maximize discrimination of spectral 
grouping in AVHRR data as well as data reduction in 
feature extraction. The result is a means to better 
understand spatial relationships through examining a 
particular phenomenon and the relationships between 
phenomena (fig. 7). 
Analysis of data from multiple spatial data layers 
requires processing techniques (algorithms) and 
procedures (processing sequences of algorithms) 
Figure 6 
suitable for manipulation of either grid cell or x,y 
coordinate structured data. The core of a GIS is 
the computer's capabilities to process spatial data 
to the specification of the user. Automated GIS 
capabilities must include: 
1) search, identification and extraction of 
spatial and non-spatial items in a data file, 2) 
format changing, 3) image data manipulation, 4) 
measurement and 5) comparison (Calkins and 
Tomlinson, 1977). 
Computer based geographic information would be of 
little use if it could not be identified and 
accessed. Data search and retrieval operations 
select user specified data for direct application or 
subsequent manipulation. These extraction 
procedures may involve entire scenes, portions of 
scenes, or specified geographic entities. Search 
and extraction functions should include browsing 
capabilities, through both graphic and non-graphic 
data, windowing capabilities, query window 
generation capabilities (i.e., calling up specific 
geographic entities) and Boolean capabilities (i.e., 
specifying extraction of data based on non-graphic 
data querying) (Dangermond, 1984). 
Manipulative functions are designed to change the 
data's physical quality to make it more useful or 
readable to the user and/or more amenable to further 
computer based spatial analysis (Calkins and 
Tomlinson, 1977). Manipulation functions include 
edge mapping, contouring, reclassification of 
polygons, map sheet manipulation (i.e., scale change 
and distortion removal), projection and coordinate 
change, coordinate rotation, and format change 
(i.e., vector to raster). 
Measurement functions enable users to inventory 
spatial data. Spatial measurements can be end 
products in themselves, or they can be utilized in 
higher levels of statistical analysis. Measurement 
capabilities consist of determining distances, 
areas, and volumes. If temporal, multi-date 
information is incorporated, then rates and 
measurement changes can be determined. 
Comparison functions allow users to detect, 
analyze and interpret relationships between 
different spatial data sets. These procedures