) showed the 
infall ranging 
onths according 
rd/or February. 
caused by the 
n and by dry, 
hse In 
son as too short 
phy over an 
evious national 
ons spread out 
hor was able to 
th a wide angle 
in discussion 
f reference. 
Super wide-angle 
ination and to 
ional coverage 
SSary. 
zed in Table B, 
«3% of the country 
en panchromatic 
, and discussed 
ical aerial 
rst author. 
nada, Central USA, 
in pan black-and- 
ssance mapping, 
iic photographs 
ial scale of about 
le photographs and 
1,1970, 1971). 
j8es could be 
be identified; 
sated; (d) meaningful 
pen forest (dry 
1 in 7 m height 
and density classes 
est (Victoria), the 
) adequate control 
scale of 1:50,000 
readily divided 
and facets). 
- 1409 - 
Recently, several colour (panchromatic) photographs at 1:140,000 of boreal 
forest in Ontario (Canada) have been examined and compared with an existing 
1959 stock-map at 1:15,840. Boreal forest was chosen, since literature over 
many years indicates that photo-interpretative techniques are readily used 
in this type of forest. 
It was observed that, in general, the photo-interpretative characteristics 
of the vegetation listed above, could be observed; but 20% crown density 
classes and meaningful parallax measurements of the images of trees and 
Stands could not be obtained although some larger tree crowns could be 
identified, Further tree species identification was very restricted although 
spruces as a group could be mapped with a much higher accuracy and reliability 
than as shown on the original forest stock map. 
Finally, several stereo-pairs of infrared colour photographs at 1:120,000 
(H: 18,460 m: 152 mm) were examined. These photographs had been taken in 
July 1972 for the US maize blight study. Appraisal of the photographs 
provided similar photo-interpretative information to the Canadian photographs; 
but include the separation of hardwoods from conifers and a non-measurable 
stereoscopic impression of stand heights. Major classes of land-use and 
several arable crops were identified against field information (e.ge alfalfa/ 
clover, pasture, maize, soya beans/oats). 
2. A Case Study (Sierra Leone) 
In order to assist the Government in the implementation of its policy on 
overall development of the agricultural sector in the country, the Ministry 
of Agriculture and Natural Resources initiated in mid-1975, a land resources 
survey project. This was financed by the United Nations Development Programme 
(UNDPY and the Government, with the FAO as executing Agency. 
The quantitative land potential evaluation considered for the initial phase 
of the project required comprehensive inventories of the rural land resources. 
It was widely accepted that such inventories can be presented conveniently in 
the form of reconnaissance maps based upon land systems and their constituent 
smaller land units and &t the sametime, up-to-date vegetation and land-use 
maps can also play an important role in agricultural planning, Consequently, 
the production of these two major types of maps was deemed necessary and action 
taken accordingly (Schwaar, 1977). 
(a) Land System Map 
The procedure retained for the reconnaissance mapping of physical factors 
required for implementing the proposed land evaluation exercise is largely 
based upon a system developed by CSIRO and others in Australia and Papua-New 
Guinea (e.g. Christian and Stewart, 1953 and 1974; Haantjens, 1969; Bleeker, 1970; 
Howard, 1970a,b; Bleeker and Birchall, 1978) and successfully applied in other 
countries of Africa (e.g. Ethiopia, Nigeria, Zambia), Near East (e.g. Jordan) and 
Far Bast (e.g. Nepal). 
The evaluation of these factors in terms of suitability for agricultural use 
was based on the latest framework for Land Evaluation developed by FAO (1976) 
adjusted to the local conditions of Sierra Leone.