Item 8(b) of the Provisional Agenda


Rome, Italy, 12-16 March 2001


Secretariat Note

Table of Contents


1. The Global Forest Resources Assessment 2000 (FRA 2000) provides a broad range of up-to-date information on the extent and condition of the world's forest resources. It has been carried out since 1951 in fulfilment of FAO's mandate to collect, analyse, interpret and disseminate information related to forestry. The FRA 2000 agenda was developed by representatives of the world's leading forestry experts in this field, during a consultation in Kotka, Finland, in 1996. In 1997, COFO and the Intergovernmental Panel on Forests (IPF IV) approved the consultation's findings and recommended that FAO serve as the lead agency for facilitating the assessment and work in partnership with other institutions in its execution.

2. FRA 2000 has compiled and analysed available information about the extent, composition, protection and utilisation of forests by country. Special attention was given to estimating the rate of change of forest resources and to documenting the various factors involved in the changes. The FRA 2000 was a transparent process in the sense that all background material and analyses were published, and each country was invited to be involved in the process. In this respect, FRA 2000 provided a participatory process for developing the most comprehensive, reliable and authoritative baseline survey of forest resources for all countries.

3. In addition to the country-by-country survey, the FRA 2000 also included an objective pan-tropical remote sensing survey of forest cover change, a set of global maps of forest cover and ecological zones, and special studies on various aspects of the interaction between people and forests. The results from FRA 2000 will be presented in printed volumes, in addition to a comprehensive version already available on the World Wide Web (

4. FRA 2000 was carried out by FAO in cooperation with donors, partners and member countries. An important partner was the UN Economic Commission for Europe, which constituted the focal point for the assessment of forests of the industrialized temperate and boreal countries.


5. FRA 2000 reported on 217 countries and other significant areas (territories and protectorates, etc.). In 2000 the world had about 3500 million hectares of forest, or 0.6 ha per caput. About half of the world's forest area was located in the tropical and sub-tropical regions, predominantly in developing countries. The other half was in the temperate and boreal regions, most of which were located in industrialized countries. Africa, Asia and North and Central America each had about 500 million hectares. Europe (including the Russian Federation) and South America had considerably more and Oceania less (see Figure 1). The four countries with the largest forest area were the Russian Federation, Brazil, Canada and the United States of America, which together accounted for half the world's forests.

6. Forest plantations made up about five percent of the world's total forest area, or almost 170 million hectares. Most of these were located in the tropical and sub-tropical regions (see Figure 2).

7. Forests covered 27 per cent of the world's total land area but their distribution among countries and regions was highly variable. Europe and South America had almost half their area under forest, whereas Africa, Asia and Oceania had less than one-fifth (Figure 3). Seventy-eight countries and other reporting areas, mainly in Asia and Africa, had less than 10 per cent of their land area forested. Seven countries (Brazil, Finland, Gabon, Indonesia, Japan, Papua New Guinea and Sweden) had forest on more than 60 per cent of their land area. French Guyana, Guyana and Suriname reported more than 90 per cent of their land under forest.

8. An objective assessment of the importance of forests to a country would entail an analysis of the benefits and availability of the various forest products and services and how they were used by that country's inhabitants. Such data were not available. However, forest area per caput could be used as a broad indicator of the potential importance of forests at the national and regional levels. For example, Asia had very little forest per caput, whereas Oceania and South America have more (see Figure 4). Only 22 countries, 18 of which were developing, had more than three hectares of forest per caput. Only about 5 per cent of the world's population lived in those countries - mostly in Brazil and the Russian Federation. Three-quarters of the world's population, on the other hand, lived in countries with less than half an hectare per caput, including most of the densely populated countries in Asia and Europe.


9. Perhaps the most sought after result of a global forest assessment is the rate of change in forest area by country. Previous FRAs have made major contributions to the world's understanding of the status of forest resources and patterns of tropical deforestation. At the same time, they have sharply heightened the debate regarding exact rates of change, methods used to capture the information and the terms and definitions used to describe forests and forest changes.

10. There are two basic aspects of forest change: afforestation and deforestation. Afforestation is the establishment of forest on a previously unforested area, and deforestation is the loss of forest. A forest, according to FAO's definition, is largely determined by the presence of tree cover and the absence of other land uses such as agriculture and grazing. Temporarily unstocked forest land (i.e. previously forested areas that are without trees for a short period of time but are expected to develop a forest cover) also qualify as forest. Change in forest area equals the difference between afforestation and deforestation. At the global level, this is usually expressed as "net deforestation", since losses of forests outweigh the gains. Gains in forest area are largely due to forest plantation establishment and natural regeneration of trees on abandoned agricultural lands. Deforestation is still largely the result of the conversion of forest to other uses such as agriculture, pasture and infrastructure.


The information base

11. At the national level, forest trend estimates for developing countries were based on national reports from recent and older forest inventories. Developing trend information from such reports constituted a major challenge for FRA 2000, as only a handful of countries carry out continuous national forest inventories with comparable time series. Over half of the countries had only one forest inventory and one-fifth of them had never carried out an inventory (see Table 1). For analyses of the vast majority of countries, therefore, the nationally provided information had to be supplemented with other sources of information.

Table 1: Forest inventories in developing countries

Continents Number of countries Number of countries
    without forest inventory partial forest inventory National Forest Inventory


Single shot2

          after 1990 before 1990







Near East







Asia & Oceania







Latin America























1) "Repeated" means systematic monitoring or inventories at a fixed interval.

2) Most countries have more than one inventory. The date (after or before 1990) refers to the most recent one.


12. Satellite images in statistically controlled time series were used to provide additional information on the status and change in forest cover in the tropics for FRA 2000. This information was relevant at the regional, ecological and pan-tropical levels, but not at national levels. Box 2 provides a description of the methodology used.

Table BOX 1: Methodologies used in the FRA 2000 assessment of tropical and sub-tropical countries

FRA's assessment of forest resources in tropical and sub-tropical countries involved two major efforts: a survey based on country information, and a regional change assessment using remote sensing.

Forest survey based on country information

To compensate for gaps in national data on forests supplied for the global assessments, FAO drew upon other sources of information to produce estimates for developing countries. FRA 2000 chose a variation of the "convergence of evidence" method, in which multiple sources of information (including inventory statistics, economic information, policy studies, etc.) were used to develop a trend line for forest change for each country. Analysts' estimates resulting from these analyses were reviewed and debated by a group of about ten forest assessment experts in FAO. The proposed final FRA 2000 change statistics were then sent to each country for their comments and approval.

Regional change assessment with remote sensing

FRA 2000 carried out a remote sensing survey to complement the one based on country information. It did this by updating information on the sample locations of the remote sensing survey carried out for FRA 1990. Statistical sampling of a uniform data source (satellite imagery) provided a comparative set of statistics. The survey sampled 10 percent of the world's tropical forests on 117 sample units to produce estimates of the state and change of these forests at regional, ecological and pan-tropical levels (not at national levels)1. Landsat satellite images provided the raw material on forest and other land-cover changes in each sample unit for the 1980-1990 and 1990-2000 periods. The principal output of the remote sensing survey was the change matrix, which illustrated and quantified how the forest and landscape changed over time. The matrices compare forest classes in 1980 with those in 1990 and 2000. The forest and land-cover classification scheme of the remote sensing survey is linked closely to the FRA forest classes for the country information and the low-resolution global forest map. In this way, FRA produces a series of information sets that are comparable and appropriate for different uses and at different scales.

Findings on forest cover change in the developing regions

13. FAO's global forest assessments of 1980 and 1990 both pointed to very high rates of deforestation at the global level, with net deforestation concentrated in the tropical regions. While deforestation was still high, FRA 2000 data available as of September 2000 indicated that the global deforestation rate for the 1990-2000 period was at least 10 percent lower than for the 1980-1990 decade. This information was provided through the analysis of 104 of the 117 sample units in the FRA 2000 remote sensing survey2. The remaining sample units will provide the definitive figures expected to be released in December 2000.

14. The survey, based on statistical sampling and carried out at regional and pan-tropical levels, showed a reduction in deforestation throughout the tropics. The decrease in deforestation was mainly observed in areas where the proportion of forest in the landscape was between 10 and 40 percent. In areas with a higher proportion of forests, the deforestation rate had not changed significantly over the past two decades. The FRA programme is now implementing independent national studies on forest change, which should help to clarify the reasons for the change in deforestation rates.

15. The situation at national level showed some significant changes over the past ten years. While many countries continued to have a high rate of deforestation, forest cover increased during the 1990s in some countries, for example India and China. One major factor influencing the forest trends was the increase of forest plantations. The reported annual planting area in developing countries was about four million hectares, of which 80 percent were in Asia. This increase was strongly linked to national policies and expanded plantation programmes.


16. The results of the assessment of temperate and boreal forests (i.e. in industrialized nations) are summarized below, with information aggregated by region. Box 2 provides a description of the methodology used for the assessment.

Table Box 2: Methodology used for the temperate and boreal assessment

Information on temperate and boreal forests was provided by national correspondents in response to a questionnaire. About 700 parameters were requested from each country, of which 70 were considered essential. The essential data were provided by practically all countries. The main sources of most of the data were national forest inventories. The correspondents adjusted national data to fit the agreed international definitions. Some estimations were necessary, but this was recorded to ensure maximum transparency. All data provided were checked and validated by the secretariat and national correspondents in an intense dialogue over nearly two years.

Changes in forest area

17. Many industrialized countries had difficulties similar to those of developing countries in estimating forest cover change during the 1990s. Change in forest area was seldom assessed directly. Data for earlier periods, which may not have been entirely comparable with new inventory statistics, frequently had to be used to calculate a rate of change. A few countries, including Australia and Canada, were not able to supply base information on forest change for FRA 2000.

18. In contrast to the high deforestation rate in many tropical and sub-tropical countries, the rate of change in forest area in most industrialized temperate and boreal countries was low. In Europe, the area of forest was expanding, while that of "other wooded land" was decreasing, for a net expansion of forest and other wooded land of 0.3 million hectares per year. There were several developments taking place simultaneously in the region:

19. In the United States of America, the forest area was also expanding while other wooded land was decreasing, for a net increase of 0.4 million hectares per year. Much of this was due to the natural transition, and reclassification, of other wooded land to forest. Most countries of the Commonwealth of Independent States (CIS) reported increases for both forest and other wooded land, giving a net increase of 1.2 million hectares per year for the region.


20. FRA 2000 was the most comprehensive, reliable and authoritative baseline survey of forest resources at the global level. It provided a broad range of up-to-date information on the world's forest resources. The assessment has stimulated important discussions concerning the status, management and protection of all forests. To ensure transparency, FRA 2000 published all background material and analyses. Results of FRA 2000 were incrementally released for the first time, using the World Wide Web as the principal instrument for publication.

21. Lessons learned from FRA 2000 and past assessments will provide the basis for the development of new and better ways of generating reliable information on the world's forests. The FRA programme will continue to seek more accurate and objective information for future global surveys, for example through increased use of sampling at the global level and capacity building in countries where there is a need to improve their forest inventories.

22. FRA 2000 results showed that many developing countries still require financial and technical support to conduct forest inventories. The scarcity of comparable multiple-date inventories illustrated the need for many of these countries to develop better mechanisms to monitor change in their forest cover. There was also a need for industrialized countries to improve their national assessments through the implementation of continuous forest inventories. Both improved and expanded surveys (i.e. beyond traditional timber inventories) will provide much of the information needed to help countries manage their forests sustainably. It is important that decision-makers are involved in the process so that the information is used and sufficient political will and financial resources are brought to bear on these efforts.

23. COFO member countries are invited to take note of the results and the lessons learned from FRA 2000, and to call for timely and high-quality inputs to support FAO's future programme in global forest resources assessment (to be considered under Agenda Item 8(d)).


1  Kotka III advised FAO to consider conducting the remote sensing survey at the global level with about 350 sample units. Financial restrictions, however, limited the work to the tropics, except for some pilot activities.

2  The FRA 2000 remote sensing survey is a statistically controlled sample, with time series that are entirely comparable. The comparison of forest changes in 1980s and 1990s is therefore highly significant on the regional and pan-tropical levels.