As the nearest alternative air photo 
mosaics, a by-product of the World Bank 
mapping projects, served as the basis for the 
photo interpretation and land use mapping. A 
major, but in this case unavoidable, 
disadvantage was the use of a relatively old 
base (4-7 years old photo mosaics in the case 
of the first 100 cities and therefore rather 
outdated). For the second series of 155 
cities, the mosaics will be somewhat more 
recent, and anyway available quite some time 
before the maps are published. 
These photo mosaics, scale 1:7500 ... 
1:12500, can be acquired without too much 
hassle in the form of bromide prints or very 
cheap blueprints. They provide an acceptable 
base for the subsequent mapping and field 
survey, and a partial source for data 
acquisition and cross-checking through photo 
interpretation. 
4 
The only way to create an up-to-date 
thematic map in the absence of recent air 
photos is updating through a field survey. The 
updating is only required in parts of the 
urban area, with a locally known location. 
Elements that are not visible on photos (e.g. 
water supply network) have to obtained from 
other sources (water company) or a field 
survey. For the largest part of the area, the 
air photo mosaics nevertheless provide the 
much needed synoptic overview and the 
essential population density and land use 
data. As long as change (particularly 
densification) within existing built-up areas 
is not very substantial, reliable data still 
can be provided. 
Planners in practice anyhow have to accept 
less than perfect data, the main issue will 
remain to get access to optimal data within 
normal constraints and to work with these 
¿■perfect data in an intelligent way. 
TANZANIA: MONITORING AND UPGRADING THE MANZESE SQUATTER AREA 
The population (about 300000 in 1967) of 
Dar es Salaam has been growing at a yearly 
rate of 6 - 9 % since the early 1960's. Most 
of the population growth has been absorbed by 
unplanned or 'squatter' areas. Manzese is the 
largest spontaneous settlement or squatter 
area, where about 34000 people were living on 
328 ha in 1967 and 96000 people in 1980 
(Sliuzas & van Vugt, 1988). The continued 
rapid growth of Manzese is taking place in the 
form of densification (expansion of existing 
houses and filling in of remaining open 
spaces) and expansion of the area (including 
settlement in 'bad lands' like flood-prone 
creek valleys). 
In a spontaneous settlement people build 
their own houses, but as the available vacant 
land is not planned for housing, 
infrastructure facilities as roads, water 
supply, sewerage, schools, clinics, etc, are 
lacking and may not be provided for many 
years. The modern policy is to upgrade the 
squatter areas by providing basic urban 
services. However, "the long term benefit of 
upgrading will depend on the continuing 
development of the area after upgrading, the 
provision made for additional future 
improvements, and the planners' ability to 
monitor and control this development” (Sliuzas 
& van Vugt, 1988) . 
In such a situation, it is essential to 
have timely access to reliable data to be able 
to monitor and analyze the developments. Only 
trustworthy data can form the basis for a 
sound analysis, both of causes and effects. 
Development scenarios then can be made, which 
enables the urban planners and administrators 
guide the development and to anticipate 
undesirable developments and counteract them. 
As in many developing countries funds are 
limited, therefore the acquisition and 
processing of data has to be inexpensive. 
Field survey methods can provide very detailed 
data, but are time-consuming and may exclude 
new developments. In the Manzese studies data 
were acquired from large format vertical 
aerial photos of 1967 and 1975, and from small 
format oblique photos made in 1980, 1981, 
1987, and 1989 (Fig.2) . The cost of oblique 
photos is limited: the major cost item is the 
rent of a light airplane, about US$ 230 per 
hour in Dar es Salaam (1989) . An area of 5 km 2 
can be covered in 1 hour from 500 m height. 
The cost of films and printing will not be 
more than US$ 100. The data take-off (photo 
interpretation, digitizing) must not be 
underestimated, but should be compared with 
the cost and quality of data from a field 
survey, including the input into the 
processing and analysis system. 
Visual interpretation of the photos 
provided data as location, type, size, and use 
of buildings, physical suitability of the 
land, vegetation cover, etc. The development 
of the area could be monitored, and 
reconstructed from old photographs, as time 
series of air photos are available. 
The data were input into a microcomputer 
based geo-information system (Usemap IV) . The 
house centroids were digitized directly from 
the air photos and, together with the house 
type, rasterized (gridcells 33.3 x 33.3 m) and 
stored in the database. The house type is 
important as a single room normally 
accommodates a household. The rooa density 
therefore is the indicator for the population 
density. A distinction can be made between 
small houses (3 rooms), large ('Swahili' type) 
houses (6 rooms) and shop houses (5 rooms). 
Moreover, houses under construction and non- 
residential buildings were classified. 
In the analysis the absorption capacity of 
the area was calculated. The land (grid cells) 
unsuitable for housing, and land used for non- 
residential purposes was first excluded. By 
applying a maximum room density norm 
(empirically derived from an analysis of 
existing densities) the total capacity could 
be derived. Subtracting the existing number of 
rooms then will result in the residual or 
absorption capacity. 
The calculated residual room capacity is 
used for estimating the maximum future 
population and facility requirements. Room 
occupancy rates and population age 
distribution can be derived from a (limited) 
field survey. Such data are required for 
calculating the demand for primary school 
rooms, clinics, etc, which can be compared 
with the actual facility provision level to 
determine the upgrading requirements (Sliuzas 
& van Vugt, 1988). The optimal locations of 
new facilities also can be analyzed using the 
geo-information system, including the 
consequences (unavoidable in an already 
densely built-up area) of demolition of 
existing houses. 
219