449 
Photo 10. 
P1 
Photo 11. 
P2 
Photo 12. 
P3 
Photo 13. 
P4 
table at a reconnaissance level. An initial 
zonation of the image provided a basic phy 
siographic structure of the region which was 
verified by fieldwork. The differences visi 
ble on the LANDSAT image relate to altitude, 
vegetation type, climate, soils and geomor 
phology. This basic structure must be under 
stood by landscape planners if they are to be 
effective, as land-use, living environment and 
community facilities will differ greatly be 
tween the plains and the mountains around 
Nairobi. 
Because the image is itself a record of the 
integration of all components in the Nairobi 
environment, the impact of man on the bio 
physical environment is visible in terms of 
agricultural practices and urban settlement. 
When mapped at the 1:1M scale, this provides 
a useful nine-zone division from which a plan 
ner can abstract areas requiring more detailed 
study. By assembling data related to an over 
view from a satellite, the larger context of 
landscape planning studies is apparent. 
The cost of LANDSAT data in relation to 
other data types is relatively small and pro 
vides a rapidly interpretable basis for fur 
ther study. Fieldwork relating to the image 
is particularly cost effective, but the 1:1M 
product is limited and larger scale products 
would probably be more useful. 
Overall the integration of all factors con 
tained in a LANDSAT image is a valuable con 
tribution to landscape planners. Extracting 
various elements from this for further study 
is a relatively easy process. 
Specific conclusions are: 
From visual analysis of a LANDSAT FCC at 
1:1M scale, a viable physiographic map can 
be compiled which can be verified on the 
ground. 
The wide availability of LANDSAT data means 
that such maps can be prepared for almost 
any location in the world, a benefit to 
those areas where aircraft data are not 
available, and in particular in developing 
countries with little planning data avail 
able . 
The physiographic map can be further elab 
orated into a land-cover map with rela 
tively little cost by fieldwork. The field 
work is made more efficient by the satel 
lite data showing location of major fea 
tures . 
The present study using one date of imagery 
only, could be improved by using data 
showing seasonal change. 
The land-cover map is a valuable step to 
wards collection of data for a detailed 
land-use map. 
Because the LANDSAT image gives an inte 
grated picture of a 185 km x 185 km area, 
the interpretation of the picture builds 
and reinforces the understanding of the 
relationship between landscape components. 
REFERENCES 
Anderson, J.R. et al. 1976. A Land-use and 
Land-cover Classification Sytem for Use 
with Remote Sensor Data. Geological Survey 
Professional Paper 964. Washington DC: 
US Government Printing Office. 
Duchhart, I. 1986. Inleiding landschapsplan- 
ning in ontwikkelingslanden. Wageningen: 
Landbouw Hogeschool. 
Government of Kenya 1983. Development Plan 
1984-1988. Nairobi: Government Printer. 
McHarg, I.L. 1969. Design with Nature. 
Philadelphia: Natural History Press. 
Tolba, M. 1982. Opening address to the Session 
of a Special Character at the 10 t Governing 
Council of the United Nations Environmental 
Program. Nairobi: UNEP.