Updated June 1998
For several years, FAO has been involved in the organisation of a project called AFRICOVER, whose goal is to establish, by and for the whole of Africa, a digital geo-referenced database on land cover and a geographic referential (geodesy, toponymy, roads, hydrography) at a 1:250,000/1:200,000 scale (1:100,000 for small countries and specific areas). A further objective of AFRICOVER is to reinforce and to build up the national and sub-regional capacities for the establishment, update and operational use of the geographic referential and land cover maps and geodatabases. This project has been prepared in response to a number of national requests for assistance to the implementation of reliable and geo-referenced information on natural resources (e.g. early warning, forest and rangeland monitoring, planning, catchment management, production of statistics, biodiversity or climate change) at sub-national, national and regional levels.
Hence, the purpose of AFRICOVER is to prepare the basic geographic information common to the information components of actual and future programmes on natural resources in African countries. From a technical point of view, the preparation of AFRICOVER products relies essentially on remote sensing data and Geographic Information Systems (GIS). The land cover will be mainly derived from visual interpretation of recent high resolution satellite images digitally enhanced. It will be done according to a homogenised and hierarchical classification system (FAO Land Cover Classification System) which conception was done by an international working group and which was finalized in a joint effort between several FAO services (SDRN/AGLS). The Eastern -African sub-region has just started the execution of the project , due to a financing of 5.4M$ US provided by the Italian government.
For the last three years, FAO has been involved in the organisation of an initiative called AFRICOVER, whose goal is to establish, by and for the whole of Africa, a digital geo-referenced database on land cover and a geographic referential (geodesy, toponymy, roads, hydrography) at a1:250,000/1:200,000 scale (1:100,000 for small countries and specific areas) using satellite remote sensing data. This base will also be generalised at a 1:1,000,000 scale, updated, made homogeneous and comparable from thematic and geographic points of view on the whole African continent.
A further objective of AFRICOVER is to reinforce and to build up the national and sub-regional capacities for the establishment, update and operational use of the geographic referential and land cover maps and geodatabases.
This project has been prepared in response to a number of national requests for assistance to the implementation of reliable and geo-referenced information on natural resources at sub-national, national and regional levels. The analysis of national needs proved that no information system on resources, concerning early warning, forest and rangeland monitoring, planning, catchment management, production of statistics, biodiversity or climate change, can do without reliable and homogeneous basic geographic information, showing both usual landmarks (infrastructures, settlement, hydrography) and land cover.
Hence, the purpose of AFRICOVER is to prepare the basic geographic information common to the information components of actual and future programmes on natural resources in African countries.
The renewable natural resources of many African countries have come under severe strain over the past two or three decades and most indicators point toward a continuation of this trend. The rate of degradation and depletion of these resources has been accelerated in proportion to the increasing population pressure. Deforestation, desertification, soil erosion and salinisation have degraded the environment so that the food security and economic development of many countries are threatened. Whilst a large amount of new remote sensing data for the assessment of natural resources is available, and technologies exist for its storage, analysis and integration, the actual situation in Africa shows a severe shortage of country-wide, as well as regional and sub-regional, quantitative and qualitative information on vegetation cover and current land use. This fact has proven to be the major limiting factor in proper planning, development and sustainable management of renewable natural resources in Africa.
At the present time, resource managers require rapid and accurate methods for accuracy and interpreting data for the development and management of the resources of Africa. earth surface remote sensing including the mapping and analysis of the spatial distribution of land cover offers considerable advantages over current alternative methods of conducting such studies. Advantages include the potential for accelerated surveys; capability to achieve a synoptic view; availability of multi-spectral data providing increased information; capability of repetitive coverage to depict seasonal and long-term changes; the relatively inexpensive cost of monitoring from space; the opportunity of integrating existing surveys into an updated monitoring system; the change detection capabilities needed by regulatory programmes for updating information on vegetation/terrain conditions; availability of imagery with minimum distortion, thereby permitting direct measurement of important agrophysical parameters; and the fact that remotely sensed data provide a permanent record. Remote sensing technology provides a vehicle for rapid collection of current detailed land use and resource data for a variety of planning purposes which is so urgently needed in Africa. In particular, high resolution satellite remote sensing can provide land cover digital data bases at a scale equal or greater than 1: 250000; which is particularly well suited for detailed environmental resource assessment.
FAO has achieved considerable progress on the pedological side through the preparation of the soils map of the world at 1:5 million scale. Comparable investment now needs to be made in the mapping of large parts of continents where a regional perspective on development are necessary but where the information resource base is least developed.
From the interpretation of this imagery and the preparation of land cover maps and their integration into an information system, a permanent record is created which lends itself to revision and updating from many sources including additional remotely sensed data, ground surveys and aerial photographs. The integration of this new source of land cover data with other resources type information will improve the quality of resource management in the region.
From a technical point of view, the preparation of AFRICOVER products relies essentially on remote sensing data and geographic information systems (GIS).The land cover will be mainly derived from visual interpretation of recent high resolution satellite images digitally enhanced. It will be done according to a homogenised and hierarchical classification system which conception is under finalization by an international working group. The geographic referential will be derived from existing topographic maps and updated from remote sensing documents and ground surveys georeferenced from GPS points. The geometrical base that will be used as a reference will depend on the quality of the geodetic network and of the topographic maps. It could be either the existing topographic maps themselves either the satellite images geocoded with GPS measurements using spatiotriangulation techniques.
In order to maximise the synergy and the scale savings, the cartographic methods will use techniques optimising the work in team and allowing task division, such as multiphase approach.
From an operational point of view, the cartography will be carried out by national teams, with the assistance of specialised international experts, originating when possible from other African countries. Several fully compatible approaches could be considered for the implementation of the project, depending on countries and sub-regions:
The FAO objective is to assist the preparation (technical, institutional and financial arrangement) and the technical monitoring of AFRICOVER. Beyond this technical assistance function, FAO offers to act as a normalisation and labelling agency for AFRICOVER: this role will consist in defining in detail, in the framework of international working groups, the standards to be applied in all African countries in terms of information, tools, analysis methods and procedures. This standardisation will greatly integrate the national and sub-regional specificities: it is absolutely necessary, from a technical point of view, to allow an homogenisation and a better distribution of final products (databases, maps), as well as important scale savings for production, update and use of georeferenced data on resources. FAO will guarantee that these standards are respected and will deliver an AFRICOVER label. FAO in some cases (as in the Eastern Africa module) is involved as execution (or co-execution) agency for the project, on request of countries and sub-regions.
The timing is estimated between 0.5 and 2 years depending on the country.
Organization potential users of the data are therefore numerous:
Practical applications of the AFRICOVER will concern all information and monitoring systems on natural resources and the environment. The AFRICOVER products can be considered as the common denominator information data set necessary to all these information systems and in particular in relation to:
Many ongoing initiatives (early warning, mapping projects, environmental information, in general or for specific environmental sectors) are present for example in the East Africa Region both at National and International level with which the project will establish suitable links in order to:
AFRICOVER was approved in July 1994 by an international consultation at the ECA headquarters, Addis Ababa, which was attended by representatives of eight African countries, 10 sub-regional organisations (IGADD, SADCC, CILSS, RCSSMRS, OACT, CRTO, CRTEAN, RECTAS, ...), 4 United Nations organisations (UNEP, UNDP, FAO, UNITAR) and 19 international and national organisations.
Today, the Eastern African sub-region (12 countries: Burundi, Democratic Republic of Congo, Djibouti, Egypt, Eritrea, Ethiopia, Kenya, Rwanda, Somalia, Sudan, Tanzania, Uganda) has already started the implementation of the project, due to a financing of 5.4M$ US provided by the Italian government. This execution will follow both a national and regional approach. For this sub-region, FAO will also act as executing agency of the project. More details on the AFRICOVER - E. Africa project are included in the section 6.
For the other sub-regions and countries, negotiations are very advanced between the different African partners and many sponsors or collaboration agencies including French Cooperation, European Union, GTZ, CIDA, World Bank. Already the REIMP World bank EU project for Central Africa adopted the Africover specifications and several EU regional programmes are considering to apply the Africover standards on more than 20 countries. In addition, a number of African countries (C™te d'Ivoire, Equatorial Guinea, Guinea, Mali, Mauritania, Mozambique, Namibia, Nigeria, Senegal, Togo, Zimbabwe) have already sent official requests; a number of requests are under way within the different ministries of other countries. Several countries such as Senegal and Tunisia have built up national AFRICOVER Working Groups which started with the formulation of project documents according to the national approach.
Technical specifications were defined through two international working groups composed of a total of 100 experts, the majority of which come from African institutions. As a result a complete set of standards and technical tools for both land cover methods and classification were defined (group 1 which met in Dakar 29-31/7/1996) and geometry and cartography (group 2 which met in Addis Ababa 23-25/9/1996). All methods were described in detail in 2 FAO publications both in French and English. A third publication is under preparation describing the mapping processing chain which is recommended for implementation in the countries. All methodological developments activities of the international working groups were supported under a FAO French trust fund. Norms were endorsed by the ECA and the African cartographic conference; they were also approved by the African department of the World bank. Many other important partners showed a high interest in the land cover classification system, such as the European Lanes network, the IGBP, the US FGDC and the Japanese Global mapping project.
In addition several pilot studies were implemented on the use of radar imagery for land cover mapping which were concluded by 2 FAO publications on "use of radar data for land cover mapping" with respectively ERS and Radarsat.
The project is implemeting the mapping of 12 countries in E. Africa many of which have amongst the lowest GDP per capita in the world. One of the strategic linkages between these countries is their association to the Nile Basin whose water availability dominates the agricultural make-up of many of these countries. The countries concerned cover an area of around 9.5 million square kilometres.
The strategic mapping of Somalia and Djibouti is included to provide continuum of coverage in Eastern Africa as well as to urgently assess the extent of the rehabilitation required with Somalia in the post-war period.
Execution will be adapted to countries' capacities as follows: in the case of the Eastern Africa module, the mapping activities are implemented in the Regional Centre for Services in Surveying Mapping and Remote Sensing (RCSSMRS) in Nairobi, where technicians from each country will be sent and trained on-the-job during the project duration and they will systematically interpret the satellite data coverages of their own countries. In this programme, the national focal points and other involved agencies in the country who are experienced in the preparation of thematic mapping and who are beneficiaries of the project will fully participate in the project implementation in addition to the provision of photointerpreters and GIS technicians, by collecting field data on a high number of GPS ground controlled points. Moreover, all national centres will receive the products of the project including the final printed thematic maps, satellite imagery and data bases.
The project main outputs are:
In particular, in order to achieve the maximum level of country involvement, it is foreseen that each Country Signatory should nominate a National Focal Point Institution (NFPI) whose aim is to help carrying out project activities at national level. Within each Country a National Working Group (NWG) will be established, which will be composed of the national bodies involved and which will reflect the multidisciplinarity of interests and represent the regional implementation structure. In cooperation with each NFPI, the NWG will, among others, assess user needs, analyze/collect ancillary data and analyse/make recommendations of final mapping Geometry/Reference issues.
An annual National Coordinators Meeting (NCM) is foreseen with the aim of discuss/make recommendations on common regional issues (e.g. strategy, methodology, data).
The project strategy will involve the utilisation of mostly Landsat Thematic Mapper (TM) 30 m spatial resolution data integrated with existing aerial photographs, maps, and of aerial surveys and field checking. Investigations are in progress to assess the possibility/availability to use other sensor data (e.g. RESURS 01, Russian high resolution imagery) to complement TM data. In particular procedures for geometric correction of TM data with Russian 2 and 10 m resolution data are under investigation.
From the technological point of view, in consideration of the needs of a standard, accurate and quick image interpretation phase together with the advances in PC, computer assisted photointerpretation, image processing, and GIS technology, an approach based an a direct image interpretation on computer screen will be adopted during the course of the project.
7. 1. Principles
In order to homogenise and take into account the different nomenclature/legend existing in the different countries, AFRICOVER E.-Africa will be the first to use the FAO Land Cover Classification system LCCS and the related SW (Ref 1,2,3) as defined by the Africover international working group on land cover in Dakar 1996. The first full operational version of the classification and software program has been developed for implementation by the Italian trust fund GCP/RAF/287/ITA Africover East Africa Project in collaboration with the FAO Soil Resources, Management and Conservation Service (AGLS).
Despite the high demand of natural resources information, many existing maps/digital data-bases are not developed to really meet multi user requirements (number of real user is often a small portion of the potential ones). One of the basic causes (generally underestimated) of such a situation is the type of classification/legend used to describe basic information as land cover and/or land use (many of these classifications/legends are generally not comparable one to another and very often single project oriented). Despite the fact that there are many classification systems in existence throughout the world, there is no single internationally accepted land cover classification system.
The FAO Land Cover Classification and its software program is a comprehensive standardised a-priori classification system designed to be able to meet specific use requirements, and created for mapping exercises, independent of the scaleor means used to map. Any identified land cover anywhere in the world can be readily accommodated (at present oceans have been excluded hence the name land cover). The proposed classification uses a set of well-defined independent diagnostic criteria which allow correlation with existing classifications and/or legends. Therefore, this system could serve as a basis for a reference system.
Land Cover Classes are defined by the combination of a set of independent diagnostic criteria, the so-called classifiers, which are hierarchically arranged to assure a high degree of geographical accuracy (also referred to as "mapability"). Because of the heterogeneity of land cover, the same set of classifiers cannot be used to define all land cover types. Also the hierarchical structure may differ from one land cover type to an other. Therefore, the classification is designed according to two main phases:
This allows the use of the most appropriate classifiers and reduces the total number of impractical combinations of classifiers. Because of the complexity of the classification and the need for standardisation, a software program of which the beta version has been developed, assist the interpretation process: it reduces heterogeneity between interpreters, and with interpretations over time. Because of the flexible manner in which the classification is set up (creation of classes at different levels of the system and the optional use of modifiers and/or environmental and/or specific technical attributes) and the tremendous number of classes possible, this innovative software program assists the user by selecting the right class going stepwise classifier by classifier. This software will be integrated into a digital image interpretation software which will allow interpretation of imagery followed by labelling of the mapping units with the land cover classes. The classification system provides a mutually exclusive Land Cover Class which comprises: (1) a unique Boolean formula (a coded string of classifiers used); (2) name/nomenclature; and (3) a unique numerical code. Both the numerical code and nomenclature can be used to build an automatically generated Legend with the created classes grouped according to the main land cover categories and their level of detail. The nomenclature can be linked to a user-defined name.
Further definition of the Land Cover Class can be achieved by adding attributes. Two types of attributes, which form separate levels in the classification, are distinguished:
From the conceptual point of view the advantages of the proposed classification are:
7.2 LCCS software program
The Land Cover Classification System software program (Figure 1.) is a stand-alone software package programmed in Access 2.0.
The program can be used to:
The proposed classification can be used as a reference base because the diagnostic criteria allow correlation with existing classifications and/or legends.