Updated December 1998
By Messrs. M. De Lannoy, D. Lantieri
67 pp, 21 figures
RSC Series No. 63, FAO, Rome 1992.
The objective of this publication is to demonstrate, on the basis of a project that took place in Chad, the possibilities inherent in the combined use of high-resolution satellite images and sample aerial photographs for making land-inventory studies. Land inventory studies are usually made in the identification and preparation phases of development projects during which agricultural or forestry activities are planned. They can also be made in the middle or at the end of a project in order to evaluate the project's achievements.
Every stage of the technique presented is illustrated by the example of the project TCP/CHD/8953 "Inventory of the Wadis of Kanem, Chad", the aim of which was to estimate the possibilities for irrigation development in a Sahelian environment.
The approach proposed consists of evaluating large regions by, in a first stage, dividing them into homogeneous areas (stratification) by means of satellite image analysis. These areas are then analysed in detail in a second stage on sample areas that are both visited on the ground and photographed from a light aircraft.
Stage 1: Stratification
General stratification and analysis
In the Chad project, the study area covered 44,000 km2 in the prefecture of Kanem. On Landsat TM colour composites at a scale of 1:100,000, it was partitioned into 17 physiographic units determined by the structure (shape, density and size) of the wadis.
In other areas, stratification might have been based on different criteria visible on the high- resolution satellite images, such as the area's land form or type of agriculture.
In practice, during the stratification stage, an analysis is also made on the satellite images of the various types of land use, since knowing these is essential to an analysis of the area. In the Chad study, the following five categories were distinguished on the image:
Characterization/substratification according to constraints
Analysis of these land-use features and the consultation of available reports on soil and water resources permitted characterization of each of the physiographic units (or strata) with six kinds of constraints to irrigation development:
Some of the physiographic units were further divided into subunits or substrata because of the variation of certain constraints within the units.
Stage 2: Collection of samples
One or two sample sites measuring 21 x 11 km were analysed in each of the strata, representing a sampling rate of 9 percent. The size and number of sites sampled can vary according to the feature to be analysed and the type of landscape. In the Chad study, aerial colour photographs were taken of the entire test area, half of the area with stereoscopic coverage. Photointerpretation permitted the production of maps of the wadis at 1:13,000 showing: the morphology (terraces, sands, lacustrine sediments, ponds, saline areas); the various types of natural vegetation (individual trees) and crops (delineated fields); as a result, areas suitable for irrigation development were defined within the wadis.
A ground mission to two or three wadis in each sample area provided confirmation and details about some of the information interpreted from the aerial photographs. In addition, analyses were made of the soils obtained in pits dug in the wadis, and laboratory analysis of water samples was made in order to better characterize the potential for irrigation of each type of wadis.
Stage 3: The final documents
The multilevel approach permits production of several types of maps, listed here by scale:
Synthetic map, at a scale of 1:1,000,000 or 1:500,000, which covers the large areas with their constraints and potentials for development. It permits the project's priorities to be quickly determined. In the Chad study, the physiographic units and subunits were outlined and regrouped into six broad types of wadis, according to the size of investment to be expected for setting up irrigation schemes.
Photomap, at 1:200,000, which covers the whole study area. it is derived directly from the satellite images. As the background for a map, it presents an imitationtrue-colour satellite image on which has been superimposed various information relating to toponymy (roads and villages), physiography (the 1 7 units mentioned above), and land use (sand encroachment, presence of crops). The photomap, or satellite image with annotations, is processed in such a way that it can be superimposed on existing topographical maps (mosaicking, geometric correction). The hues of the image are improved by digital processing for contrast enhancement.
Photomaps can be printed in hundreds of copies, They offer a useful alternative and complement to existing topographical maps because of the many kinds of updated information they provide.
Detailed maps, at: 1:13,000, which are derived from photointerpretation of the sample aerial photographs. These maps permit definition in the sample areas of the size of the area to be developed, preparation of the budget, and orientation of ground operations.
Technical and practical aspects of sampling flight
Sampling air flight is a new technique that so far has not been often used and about which very little has been published, yet especially when combined with satellite imagery. The following points are pertinent:
Aircraft: any light single- or twin-engine aircraft is recommended, as long as the owner agrees to install a camera port (10 to 15 cm in diameter) in which to lodge the camera. To secure the camera a simple lightweight base is needed, made of chipboard and shock- absorbent blocks. After the photography mission, the aircraft is very easily restored to its original condition.
Navigation device: sampling photography demands precise navigation according to a determined flight plan and an exact altitude. The Ground Position Systems (GPS) that have recently come on the market at a cost of only a few thousand dollars have proved extremely useful, even indispensable for effective air missions. During the mission, the GPS gives the aircraft's position every second in latitude, longitude and altitude, with a precision of 30 m.
Camera: the Hasselbad 6 x 6 cm camera, with a large-capacity magazine (70 exposures), a split-second release motor etc. is well-adapted to this type of photography. Kodak Ektachrome 200 ASA professional film was used. Purchase price of the complete camera and accessories is less than US $10,000.
The flight: implementation of the air mission involves some precautions and rules for adapting the camera's shutter speed (1 /500 second) and the changing of films to the speed and altitude of the aircraft. The coordination of all these factors is explained in this paper.
Economic aspects of this method
The costs and advantages of the method that combines satellite images and sample aerial photographs have been compared with those of other land-inventory methods based solely on traditional aerial photographs. It has been shown that the method discussed in this study has many advantages over traditional methods: it costs a third to a half as much, its precision is comparable or superior depending on the case, and it is about twice as quickly accomplished, The average cost of a satellite imagery and sample aerial photo study varies from US$5 to 7 per km2 - 33 percent for the satellite remote sensing component, 13 percent for the aerial photography, 42 percent for consultant staff costs, and 1 2 percent for printing of the final documents.
It is strongly recommended that the use of this new method be further developed for the preparation and monitoring of rural development projects in tropical countries.
The above publication is available from:
Viale delle Terme di Caracalla
00100 Rome, Italy