security 
1 thereby 
input data 
ccur, for 
have been 
nbol on a 
bsequently 
cy will be 
ctness and 
ion to the 
se quality 
Cy: data 
ed to data 
related to 
s that the 
he product 
indicators 
ification, 
and 80% 
? accuracy 
uracy for 
1 also be 
ng within 
to provide 
ifiable in 
or field 
is is the 
data base 
nisation’s 
standards 
ed in data 
ion. It is 
the data 
lows users 
erpret the 
d 
r fidelity 
ucture. It 
collection 
and vill 
tem, data 
cluded as 
tor data 
' content, 
tioned in 
factor, 
ification 
y rate is 
uded as a 
hin the 
attribute accuracy standards. An added 
complication is that any classification system 
used to define decay rate, will be highly 
subjective. 
A possible exception in this regard would be to 
attach a label high decay rate to those attribute 
items which can change within 6 months, but almost 
certainly will change within a year. Examples of 
these are the physical changes in shacks in slum 
areas; classification changes in type of crop 
grown on arable land; classification changes in 
the space use of buildings in city centres, etc. 
Such warnings, presented by these labels, could be 
very useful for planning purposes. 
4. CONCLUDING REMARKS 
Quite some effort is involved in setting up 
standards for data quality, but it has to be done 
and when done properly, can lead to numerous 
hidden advantages to management, such as: 
- users will have more faith in map data, since 
there is now little room for interpretation 
errors, which have lead to considerable 
financial losses in the past. A part of this 
greater faith arises from the fact that users 
are now able to evaluate for themselves whether 
the quality is sufficient for the decisions they 
have to take 
- mapping managers can use a lot of the quality 
indicators for planning purposes. Examples of 
these are: 
x the use of the summaries produced of the groups 
of features falling vithin different accuracy 
classes as indicators for map revision needs 
x if data items have a high decay rate, do not 
incorporate these into the national topo data 
base but transfer them specialised user data 
bases, where for example agriculturalists 
interested in predicting maize yield, will 
only be too happy to record all land use 
changes involving maize 
* see which conclusions can be derived from the 
accuracy tables of your organisation. For 
example, if the maps produced from the 
photogrammetric digital data collection have 
to satisfy the plotting accuracy requirement 
of being within 0.2 mm at map scale, table 4 
shows that the maximum enlargement that can be 
used from photo to map is 6.6x. 
Similarly, if the maps produced from 
digitising existing maps also have to satisfy 
the 0.2 mm accuracy requirement, then table 6 
shows that the original maps will have to be 
at a 2.7 x larger scale. 
Acknowledgement 
The authors acknowledge with thanks the assistance 
received from W.G. members in supplying the ideas 
and support in the preparation of this report. 
599