A. BLAIR RAINS*(*) A. Blair Rains, Principal Scientific Officer, Land Resources Division, Ministry of Overseas Development Tolworth Tower, Surbiton, Surrey, England.
Because of the featureless nature of many savanna areas of Africa, individual medium scale aerial photographs are difficult to interpret, whereas a synoptic view of a large area often facilitates the discrimination of different physiognomic vegetation units. It is suggested that large scale aerial photography of selected areas would, if repeated at regular intervals, provide a method of monitoring changes in the vegetation.
The investigation, mapping and monitoring of extensive areas of grassland and savanna with few obvious and easily discernible boundaries are difficult. Because of the low productivity of these areas, cost precludes frequent aerial survey. Conventional aerial survey also involves handling very large numbers of photographs (scale 1:30,000 - 1:50,000) many of which contain relatively little information. In order to obtain a synoptic view of a large area which facilitates the recognition and the delineation of boundaries between communities, blocks of several hundred photographs should be laid out, this also allows the selection of photographs for examination under the stereoscope. In this method, and when ordinary photomosaics are used, the interpreter will inevitably be distracted by tonal variations and by the edges of the individual photographs, a disadvantage not present in the imagery obtained from manned (photographic) and unmanned (M.S.S.) satellites which have been launched by the U.S.A.
The ERTS satellites (launched July 1972 and January 1975) reimage the same scene at nine-day intervals and provide a record of seasonal changes. Without minimising the value of having imagery obtained during both the dry season and the wet season, it seems unlikely that information from satellites could be used in the immediate management of grazing resources or even in the organisation of relief measures when a scarcity of fodder seems imminent.
In spite of having satellite imagery, aerial photography and field work are still required for accurate description; and monitoring of infra-red colour film at scales of 1:10,000- 1:20,000 is particularly valuable in the assessment of herbaceous vegetation if it is obtained at the end of the wet season or early in the dry season. (The only disadvantage of this reversal film is the high cost of duplicate transparencies and of colour prints.)
Because financial and practical considerations preclude ordinary surveys, it is suggested that representative areas be selected for regular monitoring by aircraft using I-R colour film and by sampling: the size of the selected areas might be between 200 and 1,000 km², and they could be regarded as large permanent quadrats or transects.